{ "checked": "2026-11-07", "fetchedAt": "2026-11-07T23:57:59.713Z", "source": { "label": "OpenRouter public model and provider listings", "url": "https://openrouter.ai/models", "api": "https://openrouter.ai/api/v1/models//endpoints" }, "note": "ernd partners with Wafer. Prices are from OpenRouter's public listings.", "units": { "prices": "USD per million tokens", "latencyS": "seconds to first token, 7-day average", "throughput": "output tokens per second, 7-day average" }, "blended": "(3 * input + output) / 4", "models": [ { "id": "deepseek/deepseek-v4.1-flash-1017", "name": "DeepSeek V4.1 Flash 1017", "author": "DeepSeek", "url": "https://ernd.ai/inference-prices/deepseek-v4-1-flash-1017", "cheapest": "Particle AI (US, Flex)", "endpoints": [ { "provider": "Particle AI (US, Flex)", "input": 0.025, "output": 0.25, "blended": 0.08125, "cacheRead": 0.0015, "context": 1048576, "quantization": null, "latencyS": 0.923, "throughput": 110, "up": true }, { "provider": "Sail Research (US)", "input": 0.101, "output": 0.15, "blended": 0.11325, "cacheRead": 0.006735, "context": 1048576, "quantization": "fp4", "latencyS": 0.71, "throughput": 72, "up": false }, { "provider": "Morph", "input": 0.022, "output": 0.4, "blended": 0.1165, "cacheRead": 0.004, "context": 1048576, "quantization": "fp8", "latencyS": 0.704, "throughput": 77, "up": true }, { "provider": "DeepInfra (Flex)", "input": 0.0784, "output": 0.2352, "blended": 0.1176, "cacheRead": 0.002352, "context": 1048576, "quantization": "fp8", "latencyS": 2.997, "throughput": 48, "up": true }, { "provider": "Relace", "input": 0.006, "output": 0.5, "blended": 0.1295, "cacheRead": 0.005, "context": 1048576, "quantization": "fp4", "latencyS": 0.683, "throughput": 116, "up": true }, { "provider": "Alibaba Cloud Int.", "input": 0.075, "output": 0.3, "blended": 0.13125, "cacheRead": 0.0075, "context": 1000000, "quantization": null, "latencyS": 0.858, "throughput": 116, "up": true }, { "provider": "DeepInfra", "input": 0.098, "output": 0.294, "blended": 0.147, "cacheRead": 0.00294, "context": 1048576, "quantization": "fp8", "latencyS": 1.182, "throughput": 58, "up": true }, { "provider": "Wafer (US)", "input": 0.038, "output": 0.5, "blended": 0.1535, "cacheRead": 0.01, "context": 1048576, "quantization": "mxfp4", "latencyS": 0.375, "throughput": 167, "up": true }, { "provider": "Particle AI (US)", "input": 0.05, "output": 0.5, "blended": 0.1625, "cacheRead": 0.003, "context": 1048576, "quantization": null, "latencyS": 1.05, "throughput": 98, "up": true }, { "provider": "Sail Research", "input": 0.1515, "output": 0.225, "blended": 0.169875, "cacheRead": 0.010103, "context": 1048576, "quantization": "fp4", "latencyS": 0.952, "throughput": 79, "up": false }, { "provider": "Baidu Qianfan", "input": 0.09772, "output": 0.39437, "blended": 0.171883, "cacheRead": 0.002094, "context": 1048576, "quantization": "fp8", "latencyS": 1.194, "throughput": 125, "up": true }, { "provider": "Wafer", "input": 0.066, "output": 0.5, "blended": 0.1745, "cacheRead": 0.0065, "context": 1048576, "quantization": "mxfp4", "latencyS": 0.371, "throughput": 165, "up": true }, { "provider": "StreamLake", "input": 0.111, "output": 0.444, "blended": 0.19425, "cacheRead": 0.00222, "context": 1024000, "quantization": "fp8", "latencyS": 1.883, "throughput": 69, "up": true }, { "provider": "Open Inference", "input": 0.03, "output": 0.7, "blended": 0.1975, "cacheRead": 0.006, "context": 1048576, "quantization": "fp4", "latencyS": 0.858, "throughput": 97, "up": true }, { "provider": "inference.net", "input": 0.08, "output": 0.55, "blended": 0.1975, "cacheRead": 0.008, "context": 1040000, "quantization": "fp8", "latencyS": 0.359, "throughput": 105, "up": true }, { "provider": "inference.net (US)", "input": 0.08, "output": 0.55, "blended": 0.1975, "cacheRead": 0.008, "context": 1040000, "quantization": "fp8", "latencyS": 0.331, "throughput": 111, "up": true }, { "provider": "Decart", "input": 0.123, "output": 0.492, "blended": 0.21525, "cacheRead": 0.00246, "context": 1048576, "quantization": "fp8", "latencyS": 1.212, "throughput": 100, "up": true }, { "provider": "DeepInfra (Priority)", "input": 0.147, "output": 0.441, "blended": 0.2205, "cacheRead": 0.00441, "context": 1048576, "quantization": "fp8", "latencyS": 1.501, "throughput": 59, "up": true }, { "provider": "Cloudflare", "input": 0.1, "output": 0.6, "blended": 0.225, "cacheRead": 0.003, "context": 1048576, "quantization": null, "latencyS": 2.163, "throughput": 90, "up": true }, { "provider": "StepFun", "input": 0.135, "output": 0.54, "blended": 0.23625, "cacheRead": 0.0027, "context": 1000000, "quantization": "fp8", "latencyS": 0.793, "throughput": 149, "up": true }, { "provider": "Ionstream", "input": 0.165, "output": 0.495, "blended": 0.2475, "cacheRead": 0.0105, "context": 1048576, "quantization": null, "latencyS": 0.895, "throughput": 119, "up": true }, { "provider": "DeepSeek", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": 0.003, "context": 1048576, "quantization": null, "latencyS": 1.089, "throughput": 91, "up": true }, { "provider": "NovitaAI", "input": 0.165, "output": 0.66, "blended": 0.28875, "cacheRead": 0.0033, "context": 1048576, "quantization": "fp8", "latencyS": 1.709, "throughput": 82, "up": true }, { "provider": "ModelRun [by Modular]", "input": 0.17, "output": 0.67, "blended": 0.295, "cacheRead": 0.006, "context": 1048576, "quantization": "fp4", "latencyS": 0.852, "throughput": 126, "up": true }, { "provider": "DekaLLM", "input": 0.18, "output": 0.68, "blended": 0.305, "cacheRead": 0.006, "context": 1048576, "quantization": null, "latencyS": 1.285, "throughput": 112, "up": true }, { "provider": "Cohere", "input": 0.195, "output": 0.66, "blended": 0.31125, "cacheRead": 0.006, "context": 1048576, "quantization": "fp8", "latencyS": 0.515, "throughput": 174, "up": true }, { "provider": "DigitalOcean", "input": 0.18, "output": 0.72, "blended": 0.315, "cacheRead": 0.006, "context": 1048576, "quantization": null, "latencyS": 0.816, "throughput": 159, "up": true }, { "provider": "GMICloud", "input": 0.18, "output": 0.72, "blended": 0.315, "cacheRead": 0.0036, "context": 1048575, "quantization": "fp8", "latencyS": 1.31, "throughput": 120, "up": true }, { "provider": "AtlasCloud", "input": 0.195, "output": 0.78, "blended": 0.34125, "cacheRead": 0.0039, "context": 1048576, "quantization": "fp8", "latencyS": 1.479, "throughput": 95, "up": true }, { "provider": "SiliconFlow", "input": 0.3, "output": 0.55, "blended": 0.3625, "cacheRead": 0.006, "context": 1048576, "quantization": "fp8", "latencyS": 0.993, "throughput": 138, "up": true }, { "provider": "Fireworks Spot", "input": 0.21, "output": 0.84, "blended": 0.3675, "cacheRead": 0.0042, "context": 1048576, "quantization": null, "latencyS": 1.134, "throughput": 121, "up": true }, { "provider": "Prime Intellect", "input": 0.1092, "output": 1.2, "blended": 0.3819, "cacheRead": 0.006, "context": 1048576, "quantization": null, "latencyS": 0.893, "throughput": 140, "up": true }, { "provider": "Reka AI", "input": 0.16, "output": 1.1, "blended": 0.395, "cacheRead": 0.008, "context": 1048576, "quantization": null, "latencyS": 1.149, "throughput": 93, "up": true }, { "provider": "Makora", "input": 0.15, "output": 1.2, "blended": 0.4125, "cacheRead": 0.006, "context": 1048576, "quantization": "fp8", "latencyS": 1.058, "throughput": 147, "up": true }, { "provider": "PrimaLabs", "input": 0.24, "output": 0.96, "blended": 0.42, "cacheRead": 0.0048, "context": 1048576, "quantization": null, "latencyS": 0.923, "throughput": 140, "up": true }, { "provider": "Modal", "input": 0.25, "output": 1, "blended": 0.4375, "cacheRead": 0.005, "context": 1048576, "quantization": null, "latencyS": 0.624, "throughput": 149, "up": true }, { "provider": "Relace (Fast)", "input": 0.3, "output": 0.9, "blended": 0.45, "cacheRead": 0.008, "context": 1048576, "quantization": "fp4", "latencyS": 0.619, "throughput": 226, "up": true }, { "provider": "Phala", "input": 0.2, "output": 1.2, "blended": 0.45, "cacheRead": 0.006, "context": 1048576, "quantization": null, "latencyS": 1.808, "throughput": 83, "up": true }, { "provider": "Crusoe", "input": 0.29, "output": 1.2, "blended": 0.5175, "cacheRead": 0.007, "context": 1048576, "quantization": "fp8", "latencyS": 0.552, "throughput": 178, "up": true }, { "provider": "Baseten", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.007, "context": 1048576, "quantization": null, "latencyS": 0.535, "throughput": 143, "up": true }, { "provider": "CoreWeave", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.006, "context": 1048576, "quantization": "fp8", "latencyS": 0.545, "throughput": 170, "up": true }, { "provider": "Fireworks", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.006, "context": 1048576, "quantization": null, "latencyS": 1.232, "throughput": 126, "up": true }, { "provider": "Mistral", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.007, "context": 1048576, "quantization": "fp8", "latencyS": 1.058, "throughput": 186, "up": true }, { "provider": "NEAR AI", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.006, "context": 1048576, "quantization": null, "latencyS": 0.67, "throughput": 43, "up": true }, { "provider": "Nebius Token Factory", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.006, "context": 1048576, "quantization": "fp8", "latencyS": 1.178, "throughput": 118, "up": true }, { "provider": "Parasail", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.006, "context": 1048576, "quantization": "fp8", "latencyS": 0.622, "throughput": 149, "up": true }, { "provider": "Together", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.006, "context": 1048576, "quantization": null, "latencyS": 0.259, "throughput": 240, "up": true }, { "provider": "Venice", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.0075, "context": 1000000, "quantization": "fp8", "latencyS": 1.151, "throughput": 120, "up": true }, { "provider": "Fireworks (US)", "input": 0.45, "output": 1.8, "blended": 0.7875, "cacheRead": 0.009, "context": 1048576, "quantization": null, "latencyS": 0.925, "throughput": 162, "up": true }, { "provider": "Baseten (Fast)", "input": 0.6, "output": 2.4, "blended": 1.05, "cacheRead": 0.014, "context": 1048576, "quantization": null, "latencyS": 0.25, "throughput": 545, "up": true }, { "provider": "Fireworks (Ultrafast)", "input": 2.4, "output": 9.6, "blended": 4.2, "cacheRead": 0.048, "context": 1048576, "quantization": null, "latencyS": null, "throughput": 213, "up": true } ] }, { "id": "deepseek/deepseek-v4.1-flash", "name": "DeepSeek V4.1 Flash", "author": "DeepSeek", "url": "https://ernd.ai/inference-prices/deepseek-v4-1-flash", "cheapest": "Baidu Qianfan", "endpoints": [ { "provider": "Baidu Qianfan", "input": 0.0507, "output": 0.2028, "blended": 0.088725, "cacheRead": 0.001014, "context": 1048576, "quantization": "fp8", "latencyS": 1.017, "throughput": 103, "up": true }, { "provider": "Morph", "input": 0.022, "output": 0.4, "blended": 0.1165, "cacheRead": 0.004, "context": 1048576, "quantization": "fp8", "latencyS": 0.712, "throughput": 73, "up": true }, { "provider": "Relace", "input": 0.006, "output": 0.5, "blended": 0.1295, "cacheRead": 0.005, "context": 1048576, "quantization": "fp4", "latencyS": 0.56, "throughput": 116, "up": true }, { "provider": "Alibaba Cloud Int.", "input": 0.075, "output": 0.3, "blended": 0.13125, "cacheRead": 0.0075, "context": 1000000, "quantization": null, "latencyS": 0.646, "throughput": 120, "up": true }, { "provider": "DeepInfra (Flex)", "input": 0.112, "output": 0.336, "blended": 0.168, "cacheRead": 0.00336, "context": 1048576, "quantization": "fp8", "latencyS": 2.156, "throughput": 64, "up": false }, { "provider": "Open Inference", "input": 0.03, "output": 0.6, "blended": 0.1725, "cacheRead": 0.006, "context": 1048576, "quantization": "fp4", "latencyS": 0.741, "throughput": 98, "up": true }, { "provider": "Wafer (US)", "input": 0.051, "output": 0.567, "blended": 0.18, "cacheRead": 0.018, "context": 1048576, "quantization": "mxfp4", "latencyS": 0.286, "throughput": 188, "up": true }, { "provider": "DeepInfra", "input": 0.14, "output": 0.42, "blended": 0.21, "cacheRead": 0.0042, "context": 1048576, "quantization": "fp8", "latencyS": 0.637, "throughput": 90, "up": true }, { "provider": "inference.net", "input": 0.103, "output": 0.6, "blended": 0.22725, "cacheRead": 0.008, "context": 1040000, "quantization": "fp8", "latencyS": 0.269, "throughput": 107, "up": true }, { "provider": "inference.net (US)", "input": 0.103, "output": 0.6, "blended": 0.22725, "cacheRead": 0.008, "context": 1040000, "quantization": "fp8", "latencyS": 0.282, "throughput": 93, "up": true }, { "provider": "AtlasCloud", "input": 0.135, "output": 0.54, "blended": 0.23625, "cacheRead": 0.0135, "context": 1048576, "quantization": "fp8", "latencyS": 0.715, "throughput": 70, "up": true }, { "provider": "Mancer", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.05, "context": 1048576, "quantization": "fp8", "latencyS": 0.744, "throughput": 58, "up": true }, { "provider": "StreamLake", "input": 0.147, "output": 0.588, "blended": 0.25725, "cacheRead": 0.00294, "context": 1024000, "quantization": "fp8", "latencyS": 1.463, "throughput": 70, "up": true }, { "provider": "CoreWeave", "input": 0.13, "output": 0.65, "blended": 0.26, "cacheRead": 0.002, "context": 1048576, "quantization": "fp8", "latencyS": 0.544, "throughput": 168, "up": true }, { "provider": "Wafer", "input": 0.19, "output": 0.65, "blended": 0.305, "cacheRead": 0.012, "context": 1048576, "quantization": "mxfp4", "latencyS": 0.266, "throughput": 189, "up": true }, { "provider": "DeepInfra (Priority)", "input": 0.21, "output": 0.63, "blended": 0.315, "cacheRead": 0.0063, "context": 1048576, "quantization": "fp8", "latencyS": 0.883, "throughput": 78, "up": true }, { "provider": "GMICloud", "input": 0.18, "output": 0.72, "blended": 0.315, "cacheRead": 0.0036, "context": 1048575, "quantization": "fp8", "latencyS": 2.113, "throughput": 59, "up": true }, { "provider": "Reka AI", "input": 0.12, "output": 0.96, "blended": 0.33, "cacheRead": 0.008, "context": 1048576, "quantization": null, "latencyS": 0.426, "throughput": 88, "up": true }, { "provider": "NovitaAI", "input": 0.195, "output": 0.78, "blended": 0.34125, "cacheRead": 0.0039, "context": 1048576, "quantization": "fp8", "latencyS": 0.962, "throughput": 99, "up": true }, { "provider": "Fireworks Spot", "input": 0.21, "output": 0.84, "blended": 0.3675, "cacheRead": 0.0042, "context": 1048576, "quantization": null, "latencyS": 0.598, "throughput": 132, "up": true }, { "provider": "Phala", "input": 0.21, "output": 0.84, "blended": 0.3675, "cacheRead": 0.0042, "context": 1048576, "quantization": null, "latencyS": 1.13, "throughput": 104, "up": true }, { "provider": "DigitalOcean", "input": 0.225, "output": 0.9, "blended": 0.39375, "cacheRead": 0.006, "context": 1048576, "quantization": null, "latencyS": 0.688, "throughput": 171, "up": true }, { "provider": "Prime Intellect", "input": 0.156, "output": 1.2, "blended": 0.417, "cacheRead": 0.0048, "context": 1048576, "quantization": null, "latencyS": 0.761, "throughput": 149, "up": true }, { "provider": "Relace (Fast)", "input": 0.3, "output": 0.9, "blended": 0.45, "cacheRead": 0.008, "context": 1048576, "quantization": "fp4", "latencyS": null, "throughput": null, "up": true }, { "provider": "Makora", "input": 0.26, "output": 1.17, "blended": 0.4875, "cacheRead": 0.0058, "context": 1048576, "quantization": "fp8", "latencyS": 0.544, "throughput": 130, "up": true }, { "provider": "Crusoe", "input": 0.29, "output": 1.2, "blended": 0.5175, "cacheRead": 0.007, "context": 1048576, "quantization": "fp8", "latencyS": 0.535, "throughput": 158, "up": true }, { "provider": "Baseten", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.007, "context": 1048576, "quantization": null, "latencyS": 0.419, "throughput": 196, "up": true }, { "provider": "Baseten (US)", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.007, "context": 1048576, "quantization": null, "latencyS": 0.336, "throughput": 197, "up": true }, { "provider": "Cohere", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.003, "context": 1048576, "quantization": "fp8", "latencyS": 0.376, "throughput": 172, "up": true }, { "provider": "Fireworks", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.006, "context": 1048576, "quantization": null, "latencyS": 0.641, "throughput": 125, "up": true }, { "provider": "Google Vertex (Global)", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.03, "context": 1048576, "quantization": null, "latencyS": 0.772, "throughput": 91, "up": true }, { "provider": "Modal", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.03, "context": 1048576, "quantization": null, "latencyS": 0.473, "throughput": 130, "up": true }, { "provider": "Nebius Token Factory", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.01, "context": 1048576, "quantization": "fp8", "latencyS": 0.852, "throughput": 112, "up": true }, { "provider": "Parasail", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.006, "context": 1048576, "quantization": "fp8", "latencyS": 0.412, "throughput": 172, "up": true }, { "provider": "SiliconFlow", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.006, "context": 1048576, "quantization": "fp8", "latencyS": 1.522, "throughput": 95, "up": true }, { "provider": "Together", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.006, "context": 1048576, "quantization": null, "latencyS": 0.288, "throughput": 217, "up": true }, { "provider": "Venice", "input": 0.3, "output": 1.2, "blended": 0.525, "cacheRead": 0.0075, "context": 1000000, "quantization": "fp8", "latencyS": 0.578, "throughput": 165, "up": true }, { "provider": "Fireworks (US)", "input": 0.45, "output": 1.8, "blended": 0.7875, "cacheRead": 0.009, "context": 1048576, "quantization": null, "latencyS": 0.368, "throughput": 200, "up": true }, { "provider": "Parasail (Fast)", "input": 0.45, "output": 1.8, "blended": 0.7875, "cacheRead": 0.009, "context": 1048576, "quantization": "fp8", "latencyS": 0.422, "throughput": 137, "up": true } ] }, { "id": "z-ai/glm-5.3-flash", "name": "GLM 5.3 Flash", "author": "Z.ai", "url": "https://ernd.ai/inference-prices/glm-5-3-flash", "cheapest": "Relace", "endpoints": [ { "provider": "Relace", "input": 0.0335, "output": 0.083, "blended": 0.045875, "cacheRead": 0.03, "context": 1048576, "quantization": "fp4", "latencyS": 0.955, "throughput": 70, "up": true }, { "provider": "Wafer (US)", "input": 0.0353, "output": 0.184, "blended": 0.072475, "cacheRead": 0.0343, "context": 1048576, "quantization": "nvfp4", "latencyS": 0.479, "throughput": 106, "up": true }, { "provider": "Wafer", "input": 0.054, "output": 0.2, "blended": 0.0905, "cacheRead": 0.053, "context": 1048576, "quantization": "nvfp4", "latencyS": 0.462, "throughput": 99, "up": true }, { "provider": "Darkbloom", "input": 0.065, "output": 0.18, "blended": 0.09375, "cacheRead": 0.012, "context": 262144, "quantization": "fp4", "latencyS": 2.562, "throughput": 29, "up": true }, { "provider": "Open Inference", "input": 0.06, "output": 0.24, "blended": 0.105, "cacheRead": 0.03, "context": 1048576, "quantization": "fp4", "latencyS": 0.78, "throughput": 111, "up": true }, { "provider": "GMICloud", "input": 0.075, "output": 0.25, "blended": 0.11875, "cacheRead": 0.015, "context": 1048576, "quantization": "fp8", "latencyS": 2.174, "throughput": 64, "up": true }, { "provider": "Tenstorrent", "input": 0.075, "output": 0.25, "blended": 0.11875, "cacheRead": 0.015, "context": 262000, "quantization": "fp8", "latencyS": 2.269, "throughput": 51, "up": true }, { "provider": "NovitaAI", "input": 0.084, "output": 0.28, "blended": 0.133, "cacheRead": 0.0168, "context": 1048576, "quantization": "fp8", "latencyS": 3.205, "throughput": 40, "up": true }, { "provider": "StreamLake", "input": 0.0885, "output": 0.295, "blended": 0.140125, "cacheRead": 0.0177, "context": 1024000, "quantization": "fp8", "latencyS": 2.833, "throughput": 40, "up": true }, { "provider": "Morph", "input": 0.122, "output": 0.248, "blended": 0.1535, "cacheRead": 0.052, "context": 1048576, "quantization": "fp8", "latencyS": 1.004, "throughput": 46, "up": true }, { "provider": "Reka AI", "input": 0.12, "output": 0.3, "blended": 0.165, "cacheRead": 0.03, "context": 1048576, "quantization": null, "latencyS": 0.682, "throughput": 79, "up": true }, { "provider": "Makora", "input": 0.14, "output": 0.24, "blended": 0.165, "cacheRead": 0.03, "context": 1048576, "quantization": "fp4", "latencyS": 0.637, "throughput": 82, "up": true }, { "provider": "inference.net", "input": 0.108, "output": 0.35, "blended": 0.1685, "cacheRead": 0.032, "context": 1048576, "quantization": "fp4", "latencyS": 0.424, "throughput": 93, "up": true }, { "provider": "inference.net (US)", "input": 0.108, "output": 0.35, "blended": 0.1685, "cacheRead": 0.032, "context": 1048576, "quantization": "fp4", "latencyS": 0.366, "throughput": 97, "up": true }, { "provider": "DigitalOcean", "input": 0.1125, "output": 0.375, "blended": 0.178125, "cacheRead": 0.03, "context": 1048576, "quantization": null, "latencyS": 0.869, "throughput": 106, "up": true }, { "provider": "Phala", "input": 0.1125, "output": 0.375, "blended": 0.178125, "cacheRead": 0.0225, "context": 1048576, "quantization": "nvfp4", "latencyS": 2.763, "throughput": 36, "up": true }, { "provider": "CoreWeave", "input": 0.12, "output": 0.38, "blended": 0.185, "cacheRead": 0.02, "context": 1048576, "quantization": "nvfp4", "latencyS": 0.683, "throughput": 123, "up": true }, { "provider": "NEAR AI", "input": 0.08, "output": 0.5, "blended": 0.185, "cacheRead": 0.012, "context": 1048576, "quantization": "fp8", "latencyS": 0.911, "throughput": 28, "up": true }, { "provider": "Particle AI", "input": 0.1, "output": 0.5, "blended": 0.2, "cacheRead": 0.03, "context": 1048576, "quantization": "nvfp4", "latencyS": 0.801, "throughput": 110, "up": true }, { "provider": "NextBit", "input": 0.12, "output": 0.45, "blended": 0.2025, "cacheRead": 0.03, "context": 1048576, "quantization": "fp4", "latencyS": 1.36, "throughput": 49, "up": true }, { "provider": "AtlasCloud", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048576, "quantization": "fp8", "latencyS": 3.281, "throughput": 37, "up": true }, { "provider": "Baseten", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048576, "quantization": "fp8", "latencyS": 0.507, "throughput": 97, "up": true }, { "provider": "Baseten (US)", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048576, "quantization": "fp8", "latencyS": 0.463, "throughput": 107, "up": true }, { "provider": "Cohere", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048576, "quantization": "nvfp4", "latencyS": 0.76, "throughput": 119, "up": true }, { "provider": "Crusoe", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048576, "quantization": "fp4", "latencyS": 0.807, "throughput": 104, "up": true }, { "provider": "DeepInfra", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048576, "quantization": "fp4", "latencyS": 0.92, "throughput": 83, "up": true }, { "provider": "Fireworks", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048576, "quantization": null, "latencyS": 0.875, "throughput": 105, "up": true }, { "provider": "Friendli", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048576, "quantization": null, "latencyS": 0.593, "throughput": 100, "up": true }, { "provider": "Mistral", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048576, "quantization": null, "latencyS": 2.183, "throughput": 61, "up": true }, { "provider": "Modal", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048576, "quantization": "fp4", "latencyS": 0.4, "throughput": 95, "up": true }, { "provider": "Nebius Token Factory", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048576, "quantization": "fp4", "latencyS": 1.102, "throughput": 80, "up": true }, { "provider": "Parasail", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048576, "quantization": "fp4", "latencyS": 0.845, "throughput": 111, "up": true }, { "provider": "Prime Intellect", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048576, "quantization": null, "latencyS": 0.948, "throughput": 93, "up": true }, { "provider": "SiliconFlow", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048576, "quantization": "fp8", "latencyS": 1.997, "throughput": 38, "up": true }, { "provider": "Together", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048575, "quantization": null, "latencyS": 1.093, "throughput": 118, "up": true }, { "provider": "Venice", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048576, "quantization": null, "latencyS": 0.893, "throughput": 105, "up": true }, { "provider": "Z.ai", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.03, "context": 1048576, "quantization": "fp8", "latencyS": 2.175, "throughput": 41, "up": true }, { "provider": "Inceptron", "input": 0.175, "output": 0.45, "blended": 0.24375, "cacheRead": 0.04, "context": 1048576, "quantization": "fp8", "latencyS": 0.595, "throughput": 90, "up": true }, { "provider": "Sail Research", "input": 0.175, "output": 0.625, "blended": 0.2875, "cacheRead": 0.0225, "context": 1048576, "quantization": "fp4", "latencyS": 0.944, "throughput": 41, "up": true }, { "provider": "Sail Research (US)", "input": 0.175, "output": 0.625, "blended": 0.2875, "cacheRead": 0.0225, "context": 1048576, "quantization": "fp4", "latencyS": 0.951, "throughput": 50, "up": true }, { "provider": "Parasail (Fast)", "input": 0.1875, "output": 0.625, "blended": 0.296875, "cacheRead": 0.0375, "context": 1048576, "quantization": "fp4", "latencyS": 1.024, "throughput": 105, "up": true } ] }, { "id": "tencent/hy4-preview", "name": "Hy4 preview", "author": "Tencent", "url": "https://ernd.ai/inference-prices/hy4-preview", "cheapest": "Tencent Cloud", "endpoints": [ { "provider": "Tencent Cloud", "input": 0.7506, "output": 2.2509, "blended": 1.125675, "cacheRead": 0.0378, "context": 1048576, "quantization": "fp8", "latencyS": 3.531, "throughput": 42, "up": true }, { "provider": "AtlasCloud", "input": 0.834, "output": 2.501, "blended": 1.25075, "cacheRead": 0.042, "context": 1048576, "quantization": "fp8", "latencyS": 2.689, "throughput": 47, "up": true }, { "provider": "Cohere", "input": 0.834, "output": 2.501, "blended": 1.25075, "cacheRead": 0.042, "context": 262144, "quantization": "mxfp4", "latencyS": 1.366, "throughput": 85, "up": true }, { "provider": "DeepInfra", "input": 0.834, "output": 2.501, "blended": 1.25075, "cacheRead": 0.042, "context": 1048576, "quantization": "fp4", "latencyS": 1.479, "throughput": 49, "up": true }, { "provider": "GMICloud", "input": 0.834, "output": 2.501, "blended": 1.25075, "cacheRead": 0.042, "context": 262144, "quantization": "bf16", "latencyS": 3.328, "throughput": 24, "up": true }, { "provider": "NovitaAI", "input": 0.834, "output": 2.501, "blended": 1.25075, "cacheRead": 0.042, "context": 1000000, "quantization": "fp8", "latencyS": 3.751, "throughput": 49, "up": true }, { "provider": "SiliconFlow", "input": 0.834, "output": 2.501, "blended": 1.25075, "cacheRead": 0.042, "context": 1048576, "quantization": "fp8", "latencyS": 1.893, "throughput": 52, "up": true } ] }, { "id": "minimax/minimax-m3.1-flash", "name": "MiniMax M3.1 Flash", "author": "MiniMax", "url": "https://ernd.ai/inference-prices/minimax-m3-1-flash", "cheapest": "MiniMax", "endpoints": [ { "provider": "MiniMax", "input": 0.075, "output": 0.3, "blended": 0.13125, "cacheRead": 0.0075, "context": 1048576, "quantization": null, "latencyS": 1.2, "throughput": 106, "up": true }, { "provider": "GMICloud", "input": 0.1125, "output": 0.45, "blended": 0.196875, "cacheRead": 0.01125, "context": 1048576, "quantization": "fp8", "latencyS": 1.944, "throughput": 93, "up": true }, { "provider": "StreamLake", "input": 0.135, "output": 0.54, "blended": 0.23625, "cacheRead": 0.0135, "context": 1000000, "quantization": "fp8", "latencyS": 2.162, "throughput": 65, "up": true }, { "provider": "DigitalOcean", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": 0.015, "context": 1048576, "quantization": null, "latencyS": 1.089, "throughput": 179, "up": true }, { "provider": "Fireworks", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": 0.015, "context": 1048576, "quantization": null, "latencyS": 0.707, "throughput": 122, "up": true }, { "provider": "inference.net", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": 0.015, "context": 1000000, "quantization": "fp4", "latencyS": 0.447, "throughput": 69, "up": true }, { "provider": "inference.net (US)", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": 0.015, "context": 1000000, "quantization": "fp4", "latencyS": 0.539, "throughput": 68, "up": true }, { "provider": "ModelRun [by Modular]", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": 0.015, "context": 1048576, "quantization": "fp4", "latencyS": 0.68, "throughput": 138, "up": true }, { "provider": "Nebius Token Factory", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": 0.015, "context": 1048576, "quantization": "fp4", "latencyS": 0.698, "throughput": 106, "up": true }, { "provider": "NovitaAI", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": 0.015, "context": 1000000, "quantization": "fp16", "latencyS": 1.485, "throughput": 86, "up": true }, { "provider": "Parasail", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": 0.015, "context": 1048576, "quantization": "fp8", "latencyS": 0.717, "throughput": 163, "up": true }, { "provider": "SambaNova", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": 0.015, "context": 1048576, "quantization": null, "latencyS": 1.467, "throughput": 54, "up": true }, { "provider": "Together", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": 0.015, "context": 1048576, "quantization": null, "latencyS": 0.545, "throughput": 118, "up": true } ] }, { "id": "deepseek/deepseek-v4-flash-0731", "name": "DeepSeek V4 Flash 0731", "author": "DeepSeek", "url": "https://ernd.ai/inference-prices/deepseek-v4-flash-0731", "cheapest": "Baidu Qianfan", "endpoints": [ { "provider": "Baidu Qianfan", "input": 0.04444, "output": 0.13332, "blended": 0.06666, "cacheRead": 0.001414, "context": 1048576, "quantization": "fp8", "latencyS": 0.858, "throughput": 96, "up": true }, { "provider": "DeepInfra", "input": 0.06, "output": 0.18, "blended": 0.09, "cacheRead": 0.015, "context": 1048576, "quantization": "fp8", "latencyS": 0.802, "throughput": 46, "up": true }, { "provider": "Wafer (US)", "input": 0.0434, "output": 0.3, "blended": 0.10755, "cacheRead": 0.035, "context": 1048576, "quantization": "nvfp4", "latencyS": 0.343, "throughput": 140, "up": true }, { "provider": "Prime Intellect", "input": 0.09, "output": 0.23, "blended": 0.125, "cacheRead": 0.029, "context": 1048576, "quantization": "fp4", "latencyS": 0.623, "throughput": 94, "up": true }, { "provider": "StreamLake", "input": 0.088, "output": 0.264, "blended": 0.132, "cacheRead": 0.0028, "context": 1024000, "quantization": "fp8", "latencyS": 1.999, "throughput": 39, "up": true }, { "provider": "Wafer (Fast)", "input": 0.1, "output": 0.23, "blended": 0.1325, "cacheRead": 0.095, "context": 1048576, "quantization": "nvfp4", "latencyS": 0.301, "throughput": 141, "up": true }, { "provider": "Reka AI", "input": 0.1, "output": 0.28, "blended": 0.145, "cacheRead": 0.04, "context": 262144, "quantization": null, "latencyS": 0.639, "throughput": 101, "up": true }, { "provider": "DigitalOcean", "input": 0.119, "output": 0.238, "blended": 0.14875, "cacheRead": 0.0238, "context": 1048576, "quantization": null, "latencyS": 0.678, "throughput": 140, "up": true }, { "provider": "Baseten (fp32)", "input": 0.13, "output": 0.26, "blended": 0.1625, "cacheRead": 0.028, "context": 1048576, "quantization": "fp32", "latencyS": 0.296, "throughput": 109, "up": true }, { "provider": "Baseten (fp8)", "input": 0.13, "output": 0.26, "blended": 0.1625, "cacheRead": 0.028, "context": 1048576, "quantization": "fp8", "latencyS": 0.328, "throughput": 137, "up": true }, { "provider": "Venice", "input": 0.13, "output": 0.26, "blended": 0.1625, "cacheRead": 0.028, "context": 1000000, "quantization": null, "latencyS": 0.587, "throughput": 114, "up": true }, { "provider": "CoreWeave", "input": 0.13, "output": 0.27, "blended": 0.165, "cacheRead": 0.05, "context": 262144, "quantization": "fp8", "latencyS": 0.416, "throughput": 103, "up": true }, { "provider": "Cohere", "input": 0.14, "output": 0.28, "blended": 0.175, "cacheRead": 0.07, "context": 1048576, "quantization": "fp8", "latencyS": 0.355, "throughput": 166, "up": true }, { "provider": "Nebius Token Factory", "input": 0.14, "output": 0.28, "blended": 0.175, "cacheRead": 0.01, "context": 1048576, "quantization": "fp8", "latencyS": 0.95, "throughput": 63, "up": true }, { "provider": "Parasail", "input": 0.14, "output": 0.28, "blended": 0.175, "cacheRead": 0.05, "context": 1048576, "quantization": "fp8", "latencyS": 0.616, "throughput": 83, "up": true }, { "provider": "Sail Research", "input": 0.15, "output": 0.3, "blended": 0.1875, "cacheRead": 0.042, "context": 1048576, "quantization": "fp4", "latencyS": 0.982, "throughput": 47, "up": true }, { "provider": "Sail Research (US)", "input": 0.15, "output": 0.3, "blended": 0.1875, "cacheRead": 0.042, "context": 1048576, "quantization": "fp4", "latencyS": 0.756, "throughput": 43, "up": true }, { "provider": "Morph", "input": 0.142, "output": 0.4, "blended": 0.2065, "cacheRead": 0.036, "context": 1048576, "quantization": "bf16", "latencyS": 0.614, "throughput": 96, "up": true }, { "provider": "Inceptron", "input": 0.13, "output": 0.495, "blended": 0.22125, "cacheRead": 0.027, "context": 1048576, "quantization": "fp4", "latencyS": 0.823, "throughput": 54, "up": true }, { "provider": "Mancer", "input": 0.175, "output": 0.5, "blended": 0.25625, "cacheRead": 0.175, "context": 1048576, "quantization": "fp8", "latencyS": 0.949, "throughput": 56, "up": true }, { "provider": "Alibaba Cloud Int.", "input": 0.176, "output": 0.528, "blended": 0.264, "cacheRead": 0.0176, "context": 1000000, "quantization": null, "latencyS": 1.859, "throughput": 96, "up": true }, { "provider": "Alibaba Cloud Int. (US-EAST-1)", "input": 0.187, "output": 0.561, "blended": 0.2805, "cacheRead": 0.0187, "context": 1000000, "quantization": null, "latencyS": 1.831, "throughput": 98, "up": true }, { "provider": "SiliconFlow", "input": 0.22, "output": 0.66, "blended": 0.33, "cacheRead": 0.028, "context": 1048576, "quantization": "fp8", "latencyS": 1.454, "throughput": 32, "up": true }, { "provider": "GMICloud", "input": 0.286, "output": 0.858, "blended": 0.429, "cacheRead": 0.0091, "context": 1048575, "quantization": "fp8", "latencyS": 2.002, "throughput": 60, "up": true }, { "provider": "Phala", "input": 0.308, "output": 0.924, "blended": 0.462, "cacheRead": 0.0196, "context": 1048576, "quantization": null, "latencyS": 1.8, "throughput": 100, "up": true }, { "provider": "NovitaAI", "input": 0.4092, "output": 1.2276, "blended": 0.6138, "cacheRead": 0.02604, "context": 1048576, "quantization": "fp8", "latencyS": 1.066, "throughput": 60, "up": true }, { "provider": "AtlasCloud", "input": 0.44, "output": 1.32, "blended": 0.66, "cacheRead": 0.028, "context": 1048576, "quantization": "fp4", "latencyS": 1.615, "throughput": 77, "up": true }, { "provider": "Chutes", "input": 0.44, "output": 1.32, "blended": 0.66, "cacheRead": 0.044, "context": 1048576, "quantization": "fp8", "latencyS": 1.765, "throughput": 64, "up": true }, { "provider": "Cloudflare", "input": 0.44, "output": 1.32, "blended": 0.66, "cacheRead": 0.014, "context": 1048576, "quantization": null, "latencyS": 1.317, "throughput": 68, "up": true } ] }, { "id": "z-ai/glm-5.3", "name": "GLM 5.3", "author": "Z.ai", "url": "https://ernd.ai/inference-prices/glm-5-3", "cheapest": "Wafer (US)", "endpoints": [ { "provider": "Wafer (US)", "input": 0.0699, "output": 1.4, "blended": 0.402425, "cacheRead": 0.0689, "context": 1048576, "quantization": "fp4", "latencyS": 0.423, "throughput": 89, "up": true }, { "provider": "Wafer", "input": 0.179, "output": 1.4, "blended": 0.48425, "cacheRead": 0.178, "context": 1048576, "quantization": "fp4", "latencyS": 0.347, "throughput": 93, "up": true }, { "provider": "Reka AI", "input": 0.14, "output": 2.1, "blended": 0.63, "cacheRead": 0.105, "context": 1048576, "quantization": null, "latencyS": 0.867, "throughput": 43, "up": true }, { "provider": "Decart", "input": 0.56, "output": 1.78, "blended": 0.865, "cacheRead": 0.1, "context": 1048576, "quantization": "fp4", "latencyS": 0.836, "throughput": 92, "up": true }, { "provider": "DekaLLM", "input": 0.18, "output": 2.95, "blended": 0.8725, "cacheRead": 0.15, "context": 1048576, "quantization": null, "latencyS": 1.256, "throughput": 49, "up": true }, { "provider": "inference.net", "input": 0.08, "output": 3.6, "blended": 0.96, "cacheRead": 0.076, "context": 1048576, "quantization": "fp4", "latencyS": 0.498, "throughput": 98, "up": true }, { "provider": "inference.net (US)", "input": 0.08, "output": 3.6, "blended": 0.96, "cacheRead": 0.076, "context": 1048576, "quantization": "fp4", "latencyS": 0.539, "throughput": 94, "up": true }, { "provider": "NovitaAI", "input": 0.63, "output": 1.98, "blended": 0.9675, "cacheRead": 0.117, "context": 1048576, "quantization": "fp8", "latencyS": 1.972, "throughput": 53, "up": true }, { "provider": "Makora", "input": 0.3, "output": 3, "blended": 0.975, "cacheRead": 0.125, "context": 1048576, "quantization": "fp4", "latencyS": 0.766, "throughput": 63, "up": true }, { "provider": "DeepInfra", "input": 0.5625, "output": 2.5, "blended": 1.046875, "cacheRead": 0.125, "context": 1048576, "quantization": "fp4", "latencyS": 1.453, "throughput": 46, "up": true }, { "provider": "Cloudflare", "input": 0.7, "output": 2.2, "blended": 1.075, "cacheRead": 0.13, "context": 1048576, "quantization": null, "latencyS": 1.815, "throughput": 65, "up": true }, { "provider": "GMICloud", "input": 0.7, "output": 2.2, "blended": 1.075, "cacheRead": 0.13, "context": 1048576, "quantization": "fp8", "latencyS": 2.672, "throughput": 50, "up": true }, { "provider": "SiliconFlow", "input": 0.7, "output": 2.2, "blended": 1.075, "cacheRead": 0.13, "context": 1048576, "quantization": "fp8", "latencyS": 2.045, "throughput": 59, "up": true }, { "provider": "Tenstorrent", "input": 0.6, "output": 2.5, "blended": 1.075, "cacheRead": 0.1, "context": 1044480, "quantization": "fp8", "latencyS": 0.77, "throughput": 75, "up": true }, { "provider": "Prime Intellect", "input": 0.5, "output": 3.08, "blended": 1.145, "cacheRead": 0.093, "context": 1048576, "quantization": null, "latencyS": 0.847, "throughput": 83, "up": true }, { "provider": "Relace", "input": 0.0703, "output": 4.3843, "blended": 1.1488, "cacheRead": 0.0703, "context": 1048576, "quantization": "fp4", "latencyS": 1.028, "throughput": 76, "up": true }, { "provider": "AkashML", "input": 0.19, "output": 4.4, "blended": 1.2425, "cacheRead": 0.19, "context": 1048576, "quantization": "fp4", "latencyS": 1.199, "throughput": 87, "up": true }, { "provider": "Phala", "input": 0.84, "output": 2.64, "blended": 1.29, "cacheRead": 0.156, "context": 1048576, "quantization": null, "latencyS": 3.236, "throughput": 38, "up": true }, { "provider": "Morph", "input": 0.74, "output": 3, "blended": 1.305, "cacheRead": 0.25, "context": 1048576, "quantization": "fp8", "latencyS": 1.552, "throughput": 32, "up": true }, { "provider": "CoreWeave", "input": 0.75, "output": 3, "blended": 1.3125, "cacheRead": 0.19, "context": 1048576, "quantization": "nvfp4", "latencyS": 0.586, "throughput": 92, "up": true }, { "provider": "Sail Research (US)", "input": 0.75, "output": 3, "blended": 1.3125, "cacheRead": 0.12, "context": 1048576, "quantization": "fp8", "latencyS": 0.98, "throughput": 72, "up": true }, { "provider": "DigitalOcean", "input": 0.91, "output": 2.86, "blended": 1.3975, "cacheRead": 0.169, "context": 1048576, "quantization": null, "latencyS": 0.545, "throughput": 88, "up": true }, { "provider": "Fireworks Spot", "input": 0.98, "output": 3.08, "blended": 1.505, "cacheRead": 0.182, "context": 1048576, "quantization": null, "latencyS": 0.646, "throughput": 88, "up": true }, { "provider": "Alibaba Cloud Int.", "input": 1.19, "output": 3.74, "blended": 1.8275, "cacheRead": 0.238, "context": 1000000, "quantization": null, "latencyS": 1.901, "throughput": 52, "up": true }, { "provider": "Sail Research", "input": 1.05, "output": 4.2, "blended": 1.8375, "cacheRead": 0.168, "context": 1048576, "quantization": "fp8", "latencyS": 0.784, "throughput": 73, "up": true }, { "provider": "Friendli", "input": 1.26, "output": 3.96, "blended": 1.935, "cacheRead": 0.234, "context": 1048576, "quantization": null, "latencyS": 0.36, "throughput": 83, "up": true }, { "provider": "Inceptron", "input": 1, "output": 4.9, "blended": 1.975, "cacheRead": 0.14, "context": 1048576, "quantization": "fp4", "latencyS": 0.733, "throughput": 100, "up": true }, { "provider": "Baidu Qianfan", "input": 1.3986, "output": 4.3956, "blended": 2.14785, "cacheRead": 0.25974, "context": 1048576, "quantization": "fp8", "latencyS": 1.058, "throughput": 61, "up": true }, { "provider": "AtlasCloud", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.26, "context": 1048576, "quantization": "fp8", "latencyS": 2.83, "throughput": 40, "up": true }, { "provider": "Baseten (fp4, US)", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.14, "context": 1048576, "quantization": "fp4", "latencyS": 2.353, "throughput": 58, "up": true }, { "provider": "Baseten (fp4)", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.14, "context": 1048576, "quantization": "fp4", "latencyS": 2.742, "throughput": 48, "up": true }, { "provider": "Crusoe", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.26, "context": 1048576, "quantization": "fp4", "latencyS": 0.466, "throughput": 118, "up": true }, { "provider": "Fireworks", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.26, "context": 1048576, "quantization": null, "latencyS": 1.309, "throughput": 77, "up": true }, { "provider": "Mistral (ZDR)", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.14, "context": 1048576, "quantization": "nvfp4", "latencyS": 0.941, "throughput": 102, "up": true }, { "provider": "Modal", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.26, "context": 1048576, "quantization": null, "latencyS": 0.337, "throughput": 112, "up": true }, { "provider": "Nebius Token Factory", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.26, "context": 1048576, "quantization": "fp4", "latencyS": 0.694, "throughput": 87, "up": true }, { "provider": "Parasail", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.26, "context": 1048576, "quantization": "fp8", "latencyS": 0.659, "throughput": 99, "up": true }, { "provider": "Together", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.26, "context": 1048575, "quantization": null, "latencyS": 0.552, "throughput": 101, "up": true }, { "provider": "Venice", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.26, "context": 1000000, "quantization": null, "latencyS": 1.035, "throughput": 72, "up": true }, { "provider": "Z.ai", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.26, "context": 1048576, "quantization": "fp8", "latencyS": 1.707, "throughput": 53, "up": true }, { "provider": "Mistral (EU)", "input": 1.54, "output": 4.84, "blended": 2.365, "cacheRead": 0.154, "context": 1048576, "quantization": "nvfp4", "latencyS": 0.911, "throughput": 132, "up": true }, { "provider": "Amazon Bedrock (US-EAST-1)", "input": 1.68, "output": 5.28, "blended": 2.58, "cacheRead": 0.312, "context": 1048576, "quantization": null, "latencyS": 1.484, "throughput": 110, "up": true }, { "provider": "Baseten (fp8, Fast)", "input": 2.1, "output": 6.6, "blended": 3.225, "cacheRead": 0.21, "context": 1048576, "quantization": "fp8", "latencyS": 0.432, "throughput": 176, "up": true }, { "provider": "Baseten (fp8, US, Fast)", "input": 2.1, "output": 6.6, "blended": 3.225, "cacheRead": 0.21, "context": 1048576, "quantization": "fp8", "latencyS": 0.443, "throughput": 179, "up": true }, { "provider": "Fireworks (Fast)", "input": 2.1, "output": 6.6, "blended": 3.225, "cacheRead": 0.39, "context": 1048576, "quantization": null, "latencyS": 1.028, "throughput": 122, "up": false }, { "provider": "Parasail (Fast)", "input": 2.1, "output": 6.6, "blended": 3.225, "cacheRead": 0.26, "context": 1048576, "quantization": "fp8", "latencyS": 0.858, "throughput": 72, "up": true }, { "provider": "Alibaba Cloud Int. (Prime)", "input": 2.8, "output": 8.8, "blended": 4.3, "cacheRead": 0.56, "context": 1000000, "quantization": null, "latencyS": 0.894, "throughput": 108, "up": true } ] }, { "id": "moonshotai/kimi-k3.1-pro", "name": "Kimi K3.1 Pro", "author": "Moonshot AI", "url": "https://ernd.ai/inference-prices/kimi-k3-1-pro", "cheapest": "Relace", "endpoints": [ { "provider": "Relace", "input": 0.4, "output": 7.8833, "blended": 2.270825, "cacheRead": 0.2, "context": 1048576, "quantization": "fp4", "latencyS": 1.326, "throughput": 65, "up": true }, { "provider": "Sail Research (US)", "input": 1.1, "output": 10, "blended": 3.325, "cacheRead": 0.2, "context": 1048576, "quantization": "fp4", "latencyS": 3.315, "throughput": 50, "up": true }, { "provider": "inference.net", "input": 1.8, "output": 9, "blended": 3.6, "cacheRead": 0.18, "context": 1048576, "quantization": "mxfp4", "latencyS": 0.961, "throughput": 72, "up": true }, { "provider": "inference.net (US)", "input": 1.8, "output": 9, "blended": 3.6, "cacheRead": 0.18, "context": 1048576, "quantization": "mxfp4", "latencyS": 1.026, "throughput": 69, "up": true }, { "provider": "PrimaLabs", "input": 1.8, "output": 9, "blended": 3.6, "cacheRead": 0.18, "context": 1048576, "quantization": "mxfp4", "latencyS": 1.621, "throughput": 70, "up": true }, { "provider": "Wafer", "input": 1.8, "output": 9, "blended": 3.6, "cacheRead": 0.18, "context": 1048576, "quantization": "mxfp4", "latencyS": 1.097, "throughput": 68, "up": true }, { "provider": "Sail Research", "input": 1.5, "output": 10, "blended": 3.625, "cacheRead": 0.2, "context": 1048576, "quantization": "fp4", "latencyS": 2.669, "throughput": 55, "up": true }, { "provider": "DeepInfra", "input": 1.9, "output": 9.5, "blended": 3.8, "cacheRead": 0.19, "context": 1048576, "quantization": "bf16", "latencyS": 11.829, "throughput": 23, "up": true }, { "provider": "AkashML", "input": 1.75, "output": 10, "blended": 3.8125, "cacheRead": 0.18, "context": 1048576, "quantization": "fp4", "latencyS": 1.789, "throughput": 63, "up": true }, { "provider": "DigitalOcean", "input": 2, "output": 10, "blended": 4, "cacheRead": 0.2, "context": 1048576, "quantization": null, "latencyS": 1.165, "throughput": 96, "up": true }, { "provider": "Fireworks", "input": 2, "output": 10, "blended": 4, "cacheRead": 0.2, "context": 1048576, "quantization": null, "latencyS": 1.839, "throughput": 50, "up": true }, { "provider": "Modal", "input": 2, "output": 10, "blended": 4, "cacheRead": 0.2, "context": 1048576, "quantization": "mxfp4", "latencyS": 0.804, "throughput": 102, "up": true }, { "provider": "Moonshot AI", "input": 2, "output": 10, "blended": 4, "cacheRead": 0.2, "context": 1048576, "quantization": "mxfp4", "latencyS": 5.573, "throughput": 25, "up": true }, { "provider": "Nebius Token Factory", "input": 2, "output": 10, "blended": 4, "cacheRead": 0.2, "context": 1048576, "quantization": "fp4", "latencyS": 2.952, "throughput": 60, "up": true }, { "provider": "Parasail", "input": 2, "output": 10, "blended": 4, "cacheRead": 0.2, "context": 1048576, "quantization": "fp4", "latencyS": 1.333, "throughput": 65, "up": true }, { "provider": "Phala", "input": 2, "output": 10, "blended": 4, "cacheRead": 0.2, "context": 1048576, "quantization": null, "latencyS": 2.715, "throughput": 51, "up": true }, { "provider": "Phala (US)", "input": 2, "output": 10, "blended": 4, "cacheRead": 0.2, "context": 1048576, "quantization": null, "latencyS": 1.406, "throughput": 61, "up": true }, { "provider": "Together", "input": 2, "output": 10, "blended": 4, "cacheRead": 0.2, "context": 1048576, "quantization": null, "latencyS": 1.215, "throughput": 49, "up": true }, { "provider": "Wafer (US)", "input": 2, "output": 10, "blended": 4, "cacheRead": 0.2, "context": 1048576, "quantization": "mxfp4", "latencyS": 0.911, "throughput": 66, "up": true }, { "provider": "Morph", "input": 1.908, "output": 12, "blended": 4.431, "cacheRead": 0.47, "context": 1048576, "quantization": "fp8", "latencyS": 1.612, "throughput": 64, "up": true }, { "provider": "Alibaba Cloud Int.", "input": 2.3, "output": 11.5, "blended": 4.6, "cacheRead": 0.23, "context": 1048576, "quantization": null, "latencyS": 2.395, "throughput": 39, "up": true }, { "provider": "inference.net (Fast)", "input": 3, "output": 15, "blended": 6, "cacheRead": 0.3, "context": 1048576, "quantization": "mxfp4", "latencyS": 0.897, "throughput": 96, "up": true }, { "provider": "inference.net (US, Fast)", "input": 3, "output": 15, "blended": 6, "cacheRead": 0.3, "context": 1048576, "quantization": "mxfp4", "latencyS": 1.241, "throughput": 84, "up": true }, { "provider": "Moonshot AI (Highspeed)", "input": 4, "output": 20, "blended": 8, "cacheRead": 0.2, "context": 1048576, "quantization": "mxfp4", "latencyS": 3.957, "throughput": 52, "up": true } ] }, { "id": "nvidia/nemotron-3-ultra-550b-a55b", "name": "Nemotron 3 Ultra", "author": "NVIDIA", "url": "https://ernd.ai/inference-prices/nemotron-3-ultra-550b-a55b", "cheapest": "DeepInfra (Flex)", "endpoints": [ { "provider": "DeepInfra (Flex)", "input": 0.4, "output": 1.76, "blended": 0.74, "cacheRead": 0.08, "context": 262144, "quantization": "fp4", "latencyS": 2.371, "throughput": 50, "up": true }, { "provider": "DeepInfra", "input": 0.5, "output": 2.2, "blended": 0.925, "cacheRead": 0.1, "context": 262144, "quantization": "fp4", "latencyS": 1.093, "throughput": 55, "up": true }, { "provider": "Baseten", "input": 0.6, "output": 2.4, "blended": 1.05, "cacheRead": 0.12, "context": 202800, "quantization": "fp4", "latencyS": 0.264, "throughput": 103, "up": true }, { "provider": "Baseten (US)", "input": 0.6, "output": 2.4, "blended": 1.05, "cacheRead": 0.12, "context": 202800, "quantization": "fp4", "latencyS": 0.274, "throughput": 108, "up": true }, { "provider": "Venice", "input": 0.625, "output": 3.125, "blended": 1.25, "cacheRead": 0.1875, "context": 256000, "quantization": "fp8", "latencyS": 1.404, "throughput": 41, "up": true } ] }, { "id": "tencent/hy3", "name": "Hy3", "author": "Tencent", "url": "https://ernd.ai/inference-prices/hy3", "cheapest": "Tencent Cloud", "endpoints": [ { "provider": "Tencent Cloud", "input": 0.0825, "output": 0.33, "blended": 0.144375, "cacheRead": 0.020625, "context": 262144, "quantization": "fp8", "latencyS": 3.059, "throughput": 44, "up": true }, { "provider": "GMICloud", "input": 0.14, "output": 0.58, "blended": 0.25, "cacheRead": 0.035, "context": 262144, "quantization": "bf16", "latencyS": 2.265, "throughput": 42, "up": true }, { "provider": "NovitaAI", "input": 0.14, "output": 0.58, "blended": 0.25, "cacheRead": 0.035, "context": 262144, "quantization": null, "latencyS": 2.771, "throughput": 51, "up": true }, { "provider": "Phala", "input": 0.15, "output": 0.64, "blended": 0.2725, "cacheRead": 0.04, "context": 262144, "quantization": null, "latencyS": 1.865, "throughput": 48, "up": true }, { "provider": "AtlasCloud", "input": 0.2, "output": 0.8, "blended": 0.35, "cacheRead": 0.05, "context": 262144, "quantization": "fp8", "latencyS": 1.618, "throughput": 62, "up": true } ] }, { "id": "nvidia/nemotron-3.5-super-vl", "name": "Nemotron 3.5 Super VL", "author": "NVIDIA", "url": "https://ernd.ai/inference-prices/nemotron-3-5-super-vl", "cheapest": "Darkbloom", "endpoints": [ { "provider": "Darkbloom", "input": 0.06, "output": 0.35, "blended": 0.1325, "cacheRead": 0.01, "context": 262144, "quantization": "nvfp4", "latencyS": 2.709, "throughput": 24, "up": true }, { "provider": "Crusoe", "input": 0.1, "output": 0.4, "blended": 0.175, "cacheRead": 0.08, "context": 1048576, "quantization": "bf16", "latencyS": 0.827, "throughput": 81, "up": true }, { "provider": "CoreWeave", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.02, "context": 1048576, "quantization": "nvfp4", "latencyS": 1.184, "throughput": 155, "up": true }, { "provider": "DeepInfra", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.02, "context": 262144, "quantization": "fp4", "latencyS": 0.638, "throughput": 220, "up": true }, { "provider": "io.net", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.02, "context": 262144, "quantization": "nvfp4", "latencyS": 1.418, "throughput": 70, "up": true }, { "provider": "Prime Intellect", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.02, "context": 1048576, "quantization": "nvfp4", "latencyS": 1.632, "throughput": 125, "up": true }, { "provider": "Together", "input": 0.15, "output": 0.5, "blended": 0.2375, "cacheRead": 0.02, "context": 262144, "quantization": null, "latencyS": 0.457, "throughput": 164, "up": true }, { "provider": "Baseten", "input": 0.2, "output": 0.5, "blended": 0.275, "cacheRead": 0.04, "context": 1048576, "quantization": "fp4", "latencyS": 0.895, "throughput": 173, "up": true }, { "provider": "Venice", "input": 0.2, "output": 0.5, "blended": 0.275, "cacheRead": 0.04, "context": 262144, "quantization": "fp4", "latencyS": 0.773, "throughput": 222, "up": true }, { "provider": "Nebius Token Factory", "input": 0.3, "output": 0.9, "blended": 0.45, "cacheRead": 0.06, "context": 1048576, "quantization": "bf16", "latencyS": 2.616, "throughput": 56, "up": true } ] }, { "id": "xiaomi/mimo-v2.6-flash", "name": "MiMo-V2.6-Flash", "author": "Xiaomi", "url": "https://ernd.ai/inference-prices/mimo-v2-6-flash", "cheapest": "GMICloud", "endpoints": [ { "provider": "GMICloud", "input": 0.112, "output": 0.224, "blended": 0.14, "cacheRead": 0.0024, "context": 1050000, "quantization": "bf16", "latencyS": 4.053, "throughput": 27, "up": true }, { "provider": "Darkbloom", "input": 0.1, "output": 0.28, "blended": 0.145, "cacheRead": 0.05, "context": 1048576, "quantization": "fp4", "latencyS": 3.301, "throughput": 21, "up": true }, { "provider": "Cohere", "input": 0.14, "output": 0.28, "blended": 0.175, "cacheRead": 0.003, "context": 1048576, "quantization": "fp8", "latencyS": 1.134, "throughput": 62, "up": true }, { "provider": "DeepInfra", "input": 0.14, "output": 0.28, "blended": 0.175, "cacheRead": 0.0028, "context": 1048576, "quantization": "fp8", "latencyS": 3.874, "throughput": 31, "up": true }, { "provider": "Makora", "input": 0.14, "output": 0.28, "blended": 0.175, "cacheRead": 0.003, "context": 1048576, "quantization": "fp8", "latencyS": 0.835, "throughput": 49, "up": true }, { "provider": "Xiaomi", "input": 0.14, "output": 0.28, "blended": 0.175, "cacheRead": 0.0028, "context": 1048576, "quantization": "fp8", "latencyS": 3.114, "throughput": 28, "up": true }, { "provider": "Venice", "input": 0.175, "output": 0.35, "blended": 0.21875, "cacheRead": 0.00375, "context": 1000000, "quantization": "fp8", "latencyS": 3.01, "throughput": 28, "up": true }, { "provider": "Tenstorrent", "input": 0.2, "output": 0.4, "blended": 0.25, "cacheRead": 0.004, "context": 1048576, "quantization": "fp8", "latencyS": 1.113, "throughput": 52, "up": true } ] }, { "id": "deepseek/deepseek-v4-flash", "name": "DeepSeek V4 Flash 0423", "author": "DeepSeek", "url": "https://ernd.ai/inference-prices/deepseek-v4-flash", "cheapest": "StreamLake", "endpoints": [ { "provider": "StreamLake", "input": 0.04354, "output": 0.08708, "blended": 0.054425, "cacheRead": 0.008708, "context": 1024000, "quantization": "fp8", "latencyS": 2.301, "throughput": 35, "up": true }, { "provider": "Baidu Qianfan", "input": 0.04424, "output": 0.08848, "blended": 0.0553, "cacheRead": 0.008848, "context": 1048576, "quantization": "fp8", "latencyS": 0.801, "throughput": 85, "up": true }, { "provider": "DeepInfra", "input": 0.09, "output": 0.18, "blended": 0.1125, "cacheRead": 0.018, "context": 1048576, "quantization": "fp8", "latencyS": 0.937, "throughput": 38, "up": true }, { "provider": "GMICloud", "input": 0.091, "output": 0.182, "blended": 0.11375, "cacheRead": 0.0182, "context": 1048575, "quantization": "fp8", "latencyS": 1.919, "throughput": 48, "up": true }, { "provider": "Venice", "input": 0.0966, "output": 0.1925, "blended": 0.120575, "cacheRead": 0.0196, "context": 1000000, "quantization": null, "latencyS": 1.051, "throughput": 44, "up": true }, { "provider": "DigitalOcean", "input": 0.098, "output": 0.196, "blended": 0.1225, "cacheRead": 0.0196, "context": 1048576, "quantization": null, "latencyS": 0.886, "throughput": 25, "up": true }, { "provider": "Fireworks", "input": 0.105, "output": 0.21, "blended": 0.13125, "cacheRead": 0.021, "context": 1048576, "quantization": null, "latencyS": 0.955, "throughput": 59, "up": false }, { "provider": "Relace", "input": 0.0413, "output": 0.45, "blended": 0.143475, "cacheRead": 0.0413, "context": 1048576, "quantization": "fp4", "latencyS": 0.552, "throughput": 107, "up": true }, { "provider": "SiliconFlow", "input": 0.13, "output": 0.28, "blended": 0.1675, "cacheRead": 0.028, "context": 1048576, "quantization": "fp8", "latencyS": 1.025, "throughput": 44, "up": true }, { "provider": "AtlasCloud", "input": 0.14, "output": 0.28, "blended": 0.175, "cacheRead": 0.028, "context": 1048576, "quantization": "fp4", "latencyS": 2.571, "throughput": 38, "up": true }, { "provider": "Parasail", "input": 0.14, "output": 0.28, "blended": 0.175, "cacheRead": 0.07, "context": 1048576, "quantization": "fp8", "latencyS": 0.684, "throughput": 67, "up": true }, { "provider": "Mancer", "input": 0.175, "output": 0.5, "blended": 0.25625, "cacheRead": 0.175, "context": 1048576, "quantization": "fp8", "latencyS": 1.262, "throughput": 30, "up": true }, { "provider": "Azure (US)", "input": 0.21, "output": 0.56, "blended": 0.2975, "cacheRead": 0.031, "context": 1048576, "quantization": null, "latencyS": 1.274, "throughput": 86, "up": true }, { "provider": "Wafer", "input": 0.6, "output": 0.17, "blended": 0.4925, "cacheRead": 0.2, "context": 1048576, "quantization": "mxfp4", "latencyS": 0.279, "throughput": 155, "up": true } ] }, { "id": "moonshotai/kimi-k3", "name": "Kimi K3", "author": "Moonshot AI", "url": "https://ernd.ai/inference-prices/kimi-k3", "cheapest": "Wafer (US)", "endpoints": [ { "provider": "Wafer (US)", "input": 0.276, "output": 9.6, "blended": 2.607, "cacheRead": 0.13, "context": 1048576, "quantization": "mxfp4", "latencyS": 0.761, "throughput": 74, "up": true }, { "provider": "Relace", "input": 0.1662, "output": 10.1398, "blended": 2.6596, "cacheRead": 0.16, "context": 1048576, "quantization": "fp4", "latencyS": 1.35, "throughput": 65, "up": true }, { "provider": "Makora", "input": 0.36, "output": 12, "blended": 3.27, "cacheRead": 0.16, "context": 1048576, "quantization": "fp4", "latencyS": 1.298, "throughput": 53, "up": true }, { "provider": "Sail Research", "input": 0.276, "output": 13.264, "blended": 3.523, "cacheRead": 0.138, "context": 1048576, "quantization": "fp4", "latencyS": 1.202, "throughput": 76, "up": true }, { "provider": "Pareto Alpha (US)", "input": 1, "output": 11.25, "blended": 3.5625, "cacheRead": 0.225, "context": 1048576, "quantization": null, "latencyS": 6.911, "throughput": 23, "up": true }, { "provider": "Sail Research (US)", "input": 0.281, "output": 13.498, "blended": 3.58525, "cacheRead": 0.14, "context": 1048576, "quantization": "fp4", "latencyS": 1.035, "throughput": 76, "up": true }, { "provider": "Chutes", "input": 1.85, "output": 9.5, "blended": 3.7625, "cacheRead": 0.185, "context": 1048576, "quantization": "mxfp4", "latencyS": 3.135, "throughput": 18, "up": true }, { "provider": "Decart", "input": 1.95, "output": 9.75, "blended": 3.9, "cacheRead": 0.195, "context": 1048576, "quantization": "mxfp4", "latencyS": 1.813, "throughput": 66, "up": true }, { "provider": "inference.net", "input": 0.234, "output": 15, "blended": 3.9255, "cacheRead": 0.195, "context": 1048576, "quantization": "fp4", "latencyS": 0.876, "throughput": 53, "up": true }, { "provider": "inference.net (US)", "input": 0.234, "output": 15, "blended": 3.9255, "cacheRead": 0.195, "context": 1048576, "quantization": "fp4", "latencyS": 0.897, "throughput": 57, "up": true }, { "provider": "Wafer (Fast)", "input": 0.155, "output": 16, "blended": 4.11625, "cacheRead": 0.154, "context": 1048576, "quantization": "mxfp4", "latencyS": 0.877, "throughput": 74, "up": true }, { "provider": "Wafer", "input": 0.192, "output": 16, "blended": 4.144, "cacheRead": 0.182, "context": 1048576, "quantization": "mxfp4", "latencyS": 0.858, "throughput": 80, "up": true }, { "provider": "Morph", "input": 2.167, "output": 12, "blended": 4.62525, "cacheRead": 0.534, "context": 1048576, "quantization": "fp8", "latencyS": 1.518, "throughput": 64, "up": true }, { "provider": "inference.net (Fast)", "input": 0.45, "output": 18, "blended": 4.8375, "cacheRead": 0.315, "context": 1048576, "quantization": "fp4", "latencyS": 0.893, "throughput": 75, "up": true }, { "provider": "inference.net (US, Fast)", "input": 0.45, "output": 18, "blended": 4.8375, "cacheRead": 0.315, "context": 1048576, "quantization": "fp4", "latencyS": 0.934, "throughput": 74, "up": true }, { "provider": "Phala", "input": 2.55, "output": 12.75, "blended": 5.1, "cacheRead": 0.255, "context": 1048576, "quantization": null, "latencyS": 4.264, "throughput": 26, "up": true }, { "provider": "DigitalOcean", "input": 2.55, "output": 12.95, "blended": 5.15, "cacheRead": 0.255, "context": 1048576, "quantization": null, "latencyS": 1.656, "throughput": 29, "up": true }, { "provider": "io.net", "input": 2.3, "output": 14, "blended": 5.225, "cacheRead": 0.23, "context": 1048320, "quantization": "fp4", "latencyS": 2.435, "throughput": 39, "up": true }, { "provider": "DeepInfra", "input": 2.85, "output": 14.25, "blended": 5.7, "cacheRead": 0.285, "context": 1048576, "quantization": "mxfp4", "latencyS": 1.155, "throughput": 38, "up": true }, { "provider": "Amazon Bedrock (US-EAST-2)", "input": 3, "output": 15, "blended": 6, "cacheRead": 0.3, "context": 1048576, "quantization": null, "latencyS": 4.054, "throughput": 27, "up": true }, { "provider": "Baseten", "input": 3, "output": 15, "blended": 6, "cacheRead": 0.3, "context": 1048576, "quantization": "fp8", "latencyS": 1.534, "throughput": 63, "up": true }, { "provider": "Fireworks", "input": 3, "output": 15, "blended": 6, "cacheRead": 0.3, "context": 1048576, "quantization": null, "latencyS": 1.793, "throughput": 48, "up": true }, { "provider": "Modal", "input": 3, "output": 15, "blended": 6, "cacheRead": 0.3, "context": 1048576, "quantization": "mxfp4", "latencyS": 0.722, "throughput": 82, "up": true }, { "provider": "Moonshot AI", "input": 3, "output": 15, "blended": 6, "cacheRead": 0.3, "context": 1048576, "quantization": null, "latencyS": 5.474, "throughput": 29, "up": true }, { "provider": "Parasail", "input": 3, "output": 15, "blended": 6, "cacheRead": 0.3, "context": 1048576, "quantization": "fp4", "latencyS": 1.458, "throughput": 47, "up": true }, { "provider": "Together", "input": 3, "output": 15, "blended": 6, "cacheRead": 0.3, "context": 1048576, "quantization": null, "latencyS": 2.725, "throughput": 33, "up": true }, { "provider": "Alibaba Cloud Int.", "input": 3.45, "output": 17.25, "blended": 6.9, "cacheRead": 0.345, "context": 1048576, "quantization": null, "latencyS": 2.174, "throughput": 30, "up": true }, { "provider": "Fireworks (Fast)", "input": 4.5, "output": 22.5, "blended": 9, "cacheRead": 0.45, "context": 1048576, "quantization": null, "latencyS": 1.339, "throughput": 76, "up": true }, { "provider": "Fireworks (US)", "input": 4.5, "output": 22.5, "blended": 9, "cacheRead": 0.45, "context": 1048576, "quantization": null, "latencyS": 1.133, "throughput": 63, "up": true }, { "provider": "Parasail (Fast)", "input": 4.5, "output": 22.5, "blended": 9, "cacheRead": 0.45, "context": 1000000, "quantization": "fp4", "latencyS": 0.546, "throughput": 106, "up": true } ] }, { "id": "nvidia/nemotron-3.5-lightning", "name": "Nemotron 3.5 Lightning", "author": "NVIDIA", "url": "https://ernd.ai/inference-prices/nemotron-3-5-lightning", "cheapest": "Darkbloom", "endpoints": [ { "provider": "Darkbloom", "input": 0.025, "output": 0.12, "blended": 0.04875, "cacheRead": 0.0125, "context": 262144, "quantization": "int4", "latencyS": 1.143, "throughput": 74, "up": true }, { "provider": "io.net", "input": 0.0469, "output": 0.134, "blended": 0.068675, "cacheRead": 0.02345, "context": 262144, "quantization": null, "latencyS": 0.57, "throughput": 106, "up": true }, { "provider": "DeepInfra", "input": 0.06, "output": 0.16, "blended": 0.085, "cacheRead": 0.03, "context": 262144, "quantization": "bf16", "latencyS": 0.413, "throughput": 146, "up": true }, { "provider": "Fireworks", "input": 0.05, "output": 0.2, "blended": 0.0875, "cacheRead": 0.01, "context": 262144, "quantization": null, "latencyS": 0.282, "throughput": 327, "up": true }, { "provider": "CoreWeave", "input": 0.07, "output": 0.2, "blended": 0.1025, "cacheRead": 0.04, "context": 262144, "quantization": "bf16", "latencyS": 0.248, "throughput": 164, "up": true } ] }, { "id": "xiaomi/mimo-v2.6-pro", "name": "MiMo-V2.6-Pro", "author": "Xiaomi", "url": "https://ernd.ai/inference-prices/mimo-v2-6-pro", "cheapest": "DeepInfra", "endpoints": [ { "provider": "GMICloud", "input": 0.32625, "output": 0.6525, "blended": 0.407813, "cacheRead": 0.003, "context": 1050000, "quantization": "bf16", "latencyS": 3.5, "throughput": 30, "up": false }, { "provider": "DeepInfra", "input": 0.387, "output": 0.783, "blended": 0.486, "cacheRead": 0.00324, "context": 1048576, "quantization": "fp8", "latencyS": 1.167, "throughput": 28, "up": true }, { "provider": "Xiaomi", "input": 0.435, "output": 0.87, "blended": 0.54375, "cacheRead": 0.0036, "context": 1048576, "quantization": "fp8", "latencyS": 2.817, "throughput": 28, "up": true }, { "provider": "StreamLake", "input": 0.522, "output": 1.044, "blended": 0.6525, "cacheRead": 0.00432, "context": 1000000, "quantization": null, "latencyS": 1.669, "throughput": 32, "up": true }, { "provider": "PrimaLabs", "input": 4, "output": 8, "blended": 5, "cacheRead": 0.036, "context": 1048576, "quantization": "fp8", "latencyS": 0.844, "throughput": 132, "up": true } ] }, { "id": "z-ai/glm-5.2", "name": "GLM 5.2", "author": "Z.ai", "url": "https://ernd.ai/inference-prices/glm-5-2", "cheapest": "Baidu Qianfan (fp8)", "endpoints": [ { "provider": "Baidu Qianfan (fp8)", "input": 0.189, "output": 0.594, "blended": 0.29025, "cacheRead": 0.0351, "context": 1048576, "quantization": "fp8", "latencyS": 0.947, "throughput": 65, "up": true }, { "provider": "StreamLake", "input": 0.28, "output": 0.88, "blended": 0.43, "cacheRead": 0.052, "context": 1024000, "quantization": "fp8", "latencyS": 1.495, "throughput": 45, "up": true }, { "provider": "DeepInfra (Flex)", "input": 0.45, "output": 1.44, "blended": 0.6975, "cacheRead": 0.084, "context": 1048576, "quantization": "fp4", "latencyS": 3.039, "throughput": 37, "up": true }, { "provider": "DeepInfra", "input": 0.5625, "output": 1.8, "blended": 0.871875, "cacheRead": 0.105, "context": 1048576, "quantization": "fp4", "latencyS": 1.398, "throughput": 40, "up": true }, { "provider": "Ambient", "input": 0.6, "output": 2, "blended": 0.95, "cacheRead": 0.15, "context": 202752, "quantization": "fp8", "latencyS": 0.993, "throughput": 76, "up": true }, { "provider": "NovitaAI", "input": 0.6496, "output": 2.0416, "blended": 0.9976, "cacheRead": 0.12064, "context": 1048576, "quantization": "fp8", "latencyS": 2.782, "throughput": 40, "up": true }, { "provider": "CoreWeave", "input": 0.62, "output": 2.42, "blended": 1.07, "cacheRead": 0.14, "context": 1048576, "quantization": "fp4", "latencyS": 0.536, "throughput": 145, "up": true }, { "provider": "DigitalOcean", "input": 0.7, "output": 2.2, "blended": 1.075, "cacheRead": 0.105, "context": 1048576, "quantization": null, "latencyS": 0.97, "throughput": 73, "up": true }, { "provider": "SiliconFlow", "input": 0.7, "output": 2.2, "blended": 1.075, "cacheRead": 0.13, "context": 1048576, "quantization": "fp8", "latencyS": 1.085, "throughput": 80, "up": true }, { "provider": "Morph", "input": 0.74, "output": 3, "blended": 1.305, "cacheRead": 0.25, "context": 1048576, "quantization": "fp8", "latencyS": 4.638, "throughput": 32, "up": true }, { "provider": "AtlasCloud", "input": 0.938, "output": 2.948, "blended": 1.4405, "cacheRead": 0.1742, "context": 1048576, "quantization": "fp8", "latencyS": 1.219, "throughput": 50, "up": true }, { "provider": "Alibaba Cloud Int.", "input": 0.966, "output": 3.036, "blended": 1.4835, "cacheRead": 0.1932, "context": 1000000, "quantization": "fp8", "latencyS": 1.024, "throughput": 67, "up": true }, { "provider": "Wafer (US)", "input": 0.006, "output": 6.8, "blended": 1.7045, "cacheRead": 0.005, "context": 1048576, "quantization": "fp4", "latencyS": 0.461, "throughput": 116, "up": true }, { "provider": "Alibaba Cloud Int. (US-EAST-1)", "input": 1.12, "output": 3.52, "blended": 1.72, "cacheRead": 0.224, "context": 1000000, "quantization": "fp8", "latencyS": 1.504, "throughput": 54, "up": true }, { "provider": "Relace", "input": 0.0065, "output": 7.9057, "blended": 1.9813, "cacheRead": 0.0065, "context": 1048576, "quantization": "fp4", "latencyS": 1.368, "throughput": 49, "up": true }, { "provider": "Cloudflare", "input": 0.19, "output": 8, "blended": 2.1425, "cacheRead": 0.19, "context": 262144, "quantization": null, "latencyS": 0.982, "throughput": 85, "up": true }, { "provider": "Wafer", "input": 0.19, "output": 8, "blended": 2.1425, "cacheRead": 0.18, "context": 1048576, "quantization": "fp4", "latencyS": 0.417, "throughput": 104, "up": true }, { "provider": "Baseten", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.14, "context": 1048576, "quantization": "fp8", "latencyS": 2.409, "throughput": 51, "up": true }, { "provider": "Baseten (US)", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.14, "context": 1048576, "quantization": "fp8", "latencyS": 2.523, "throughput": 55, "up": true }, { "provider": "Friendli", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.26, "context": 1048576, "quantization": null, "latencyS": 0.434, "throughput": 93, "up": true }, { "provider": "GMICloud", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.26, "context": 1048576, "quantization": "fp8", "latencyS": 2.875, "throughput": 33, "up": true }, { "provider": "Inceptron", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.19, "context": 1048576, "quantization": "fp4", "latencyS": 0.531, "throughput": 106, "up": true }, { "provider": "Mistral (nvfp4, ZDR)", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.14, "context": 1048576, "quantization": "nvfp4", "latencyS": 1.271, "throughput": 77, "up": true }, { "provider": "Nebius Token Factory", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.14, "context": 1048576, "quantization": "fp4", "latencyS": 1.758, "throughput": 45, "up": true }, { "provider": "Parasail", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.26, "context": 262144, "quantization": "fp4", "latencyS": 0.868, "throughput": 105, "up": true }, { "provider": "Together", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.26, "context": 1048575, "quantization": null, "latencyS": 0.362, "throughput": 101, "up": true }, { "provider": "Venice", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.26, "context": 1000000, "quantization": "fp8", "latencyS": 2.386, "throughput": 78, "up": true }, { "provider": "Z.ai", "input": 1.4, "output": 4.4, "blended": 2.15, "cacheRead": 0.26, "context": 1048576, "quantization": "fp8", "latencyS": 1.418, "throughput": 54, "up": true }, { "provider": "Mistral (EU)", "input": 1.54, "output": 4.84, "blended": 2.365, "cacheRead": 0.154, "context": 1048576, "quantization": null, "latencyS": 0.794, "throughput": 97, "up": true }, { "provider": "Baseten (Fast)", "input": 2.1, "output": 6.6, "blended": 3.225, "cacheRead": 0.21, "context": 1048576, "quantization": "fp8", "latencyS": 0.388, "throughput": 194, "up": true }, { "provider": "Baseten (US, Fast)", "input": 2.1, "output": 6.6, "blended": 3.225, "cacheRead": 0.21, "context": 1048576, "quantization": "fp8", "latencyS": 0.403, "throughput": 194, "up": true }, { "provider": "Alibaba Cloud Int. (Fast)", "input": 2.31, "output": 7.26, "blended": 3.5475, "cacheRead": 0.462, "context": 1000000, "quantization": "fp8", "latencyS": 1.051, "throughput": 79, "up": true }, { "provider": "Baidu Qianfan (fp4, Fast)", "input": 2.25, "output": 7.88, "blended": 3.6575, "cacheRead": 0.56, "context": 1048576, "quantization": "fp4", "latencyS": 1.021, "throughput": 69, "up": true } ] }, { "id": "deepseek/deepseek-v4-pro-0813", "name": "DeepSeek V4 Pro 0813", "author": "DeepSeek", "url": "https://ernd.ai/inference-prices/deepseek-v4-pro-0813", "cheapest": "Alibaba Cloud Int.", "endpoints": [ { "provider": "Alibaba Cloud Int.", "input": 0.5808, "output": 1.7424, "blended": 0.8712, "cacheRead": 0.05808, "context": 1000000, "quantization": null, "latencyS": 1.139, "throughput": 26, "up": true }, { "provider": "Baidu Qianfan", "input": 0.65868, "output": 1.97604, "blended": 0.98802, "cacheRead": 0.020958, "context": 1048576, "quantization": "fp8", "latencyS": 1.316, "throughput": 48, "up": true }, { "provider": "Alibaba Cloud Int. (US-EAST-1)", "input": 0.66, "output": 1.98, "blended": 0.99, "cacheRead": 0.066, "context": 1000000, "quantization": null, "latencyS": 1.175, "throughput": 23, "up": true }, { "provider": "DeepSeek", "input": 0.66, "output": 1.98, "blended": 0.99, "cacheRead": 0.022, "context": 1048576, "quantization": null, "latencyS": 1.195, "throughput": 52, "up": true }, { "provider": "StreamLake", "input": 0.66, "output": 1.98, "blended": 0.99, "cacheRead": 0.022, "context": 1024000, "quantization": null, "latencyS": 1.727, "throughput": 43, "up": true }, { "provider": "Wafer (US)", "input": 0.111, "output": 4.44, "blended": 1.19325, "cacheRead": 0.106, "context": 1048576, "quantization": "mxfp4", "latencyS": 0.284, "throughput": 86, "up": true }, { "provider": "Wafer", "input": 0.285, "output": 4.3, "blended": 1.28875, "cacheRead": 0.284, "context": 1048576, "quantization": "mxfp4", "latencyS": 0.27, "throughput": 79, "up": true }, { "provider": "Phala", "input": 0.957, "output": 2.8776, "blended": 1.43715, "cacheRead": 0.099, "context": 1048576, "quantization": null, "latencyS": 2.714, "throughput": 52, "up": true }, { "provider": "NovitaAI", "input": 0.99, "output": 2.97, "blended": 1.485, "cacheRead": 0.033, "context": 1048576, "quantization": "fp8", "latencyS": 1.621, "throughput": 45, "up": true }, { "provider": "GMICloud", "input": 1.056, "output": 3.168, "blended": 1.584, "cacheRead": 0.0352, "context": 1048575, "quantization": "fp8", "latencyS": 2.472, "throughput": 45, "up": true }, { "provider": "Sail Research", "input": 1.35, "output": 2.4, "blended": 1.6125, "cacheRead": 0.195, "context": 1048576, "quantization": "fp4", "latencyS": 0.397, "throughput": 65, "up": true }, { "provider": "Sail Research (US)", "input": 1.35, "output": 2.4, "blended": 1.6125, "cacheRead": 0.195, "context": 1048576, "quantization": "fp4", "latencyS": 0.409, "throughput": 66, "up": true }, { "provider": "DeepInfra", "input": 1.3, "output": 2.6, "blended": 1.625, "cacheRead": 0.1, "context": 1048576, "quantization": "fp8", "latencyS": 1.085, "throughput": 41, "up": true }, { "provider": "CoreWeave", "input": 1.31, "output": 3.96, "blended": 1.9725, "cacheRead": 0.044, "context": 1048576, "quantization": "fp8", "latencyS": 0.59, "throughput": 104, "up": true }, { "provider": "AtlasCloud", "input": 1.32, "output": 3.96, "blended": 1.98, "cacheRead": 0.132, "context": 1048576, "quantization": "fp8", "latencyS": 1.377, "throughput": 28, "up": true }, { "provider": "Baseten", "input": 1.32, "output": 3.96, "blended": 1.98, "cacheRead": 0.132, "context": 1048576, "quantization": "fp4", "latencyS": 0.773, "throughput": 45, "up": true }, { "provider": "Cloudflare", "input": 1.32, "output": 3.96, "blended": 1.98, "cacheRead": 0.044, "context": 1048576, "quantization": null, "latencyS": 0.803, "throughput": 32, "up": true }, { "provider": "DigitalOcean", "input": 1.32, "output": 3.96, "blended": 1.98, "cacheRead": 0.044, "context": 1048576, "quantization": null, "latencyS": 1.024, "throughput": 83, "up": true }, { "provider": "Nebius Token Factory", "input": 1.32, "output": 3.96, "blended": 1.98, "cacheRead": 0.04, "context": 1048576, "quantization": "fp8", "latencyS": 0.483, "throughput": 66, "up": true }, { "provider": "Parasail", "input": 1.32, "output": 3.96, "blended": 1.98, "cacheRead": 0.044, "context": 1048576, "quantization": "fp8", "latencyS": 0.593, "throughput": 35, "up": true }, { "provider": "SiliconFlow", "input": 1.32, "output": 3.96, "blended": 1.98, "cacheRead": 0.044, "context": 1048576, "quantization": "fp8", "latencyS": 1.277, "throughput": 25, "up": true }, { "provider": "Together", "input": 1.32, "output": 3.96, "blended": 1.98, "cacheRead": 0.13, "context": 1048576, "quantization": null, "latencyS": 0.523, "throughput": 89, "up": true }, { "provider": "Relace", "input": 0.0999, "output": 8, "blended": 2.074925, "cacheRead": 0.0999, "context": 1048576, "quantization": "fp4", "latencyS": 0.674, "throughput": 20, "up": true }, { "provider": "Venice", "input": 1.65, "output": 4.95, "blended": 2.475, "cacheRead": 0.165, "context": 1000000, "quantization": null, "latencyS": 1.213, "throughput": 26, "up": true }, { "provider": "Parasail (Fast)", "input": 1.98, "output": 5.94, "blended": 2.97, "cacheRead": 0.066, "context": 1048576, "quantization": "fp8", "latencyS": 0.389, "throughput": 67, "up": true } ] }, { "id": "qwen/qwen3.8-27b", "name": "Qwen3.8 27B", "author": "Qwen", "url": "https://ernd.ai/inference-prices/qwen3-8-27b", "cheapest": "Chutes", "endpoints": [ { "provider": "Chutes", "input": 0.039, "output": 1.25, "blended": 0.34175, "cacheRead": 0.0039, "context": 262144, "quantization": "fp8", "latencyS": 2.513, "throughput": 28, "up": true }, { "provider": "NEAR AI", "input": 0.04, "output": 1.35, "blended": 0.3675, "cacheRead": 0.018, "context": 262144, "quantization": "fp8", "latencyS": 1.259, "throughput": 65, "up": true }, { "provider": "StreamLake", "input": 0.175, "output": 1.05, "blended": 0.39375, "cacheRead": 0.035, "context": 1024000, "quantization": null, "latencyS": 5.24, "throughput": 54, "up": true }, { "provider": "Darkbloom", "input": 0.035, "output": 1.5, "blended": 0.40125, "cacheRead": 0.0175, "context": 262144, "quantization": "fp4", "latencyS": 1.366, "throughput": 52, "up": true }, { "provider": "Prime Intellect", "input": 0.034, "output": 1.864, "blended": 0.4915, "cacheRead": 0.031, "context": 131072, "quantization": "fp8", "latencyS": 0.851, "throughput": 21, "up": true }, { "provider": "Ionstream", "input": 0.04, "output": 1.95, "blended": 0.5175, "cacheRead": 0.02, "context": 262144, "quantization": "fp8", "latencyS": 0.988, "throughput": 46, "up": true }, { "provider": "DeepInfra", "input": 0.15, "output": 1.875, "blended": 0.58125, "cacheRead": 0.0375, "context": 262144, "quantization": "bf16", "latencyS": 0.702, "throughput": 34, "up": true }, { "provider": "Modal", "input": 0.04, "output": 2.3, "blended": 0.605, "cacheRead": 0.02, "context": 262144, "quantization": null, "latencyS": 0.958, "throughput": 113, "up": true }, { "provider": "Tenstorrent", "input": 0.15, "output": 2, "blended": 0.6125, "cacheRead": 0.05, "context": 262144, "quantization": "fp8", "latencyS": 0.608, "throughput": 81, "up": true }, { "provider": "Wafer", "input": 0.06, "output": 2.3, "blended": 0.62, "cacheRead": 0.059, "context": 262144, "quantization": "fp8", "latencyS": 0.8, "throughput": 83, "up": true }, { "provider": "DekaLLM", "input": 0.04, "output": 2.5, "blended": 0.655, "cacheRead": 0.02, "context": 262144, "quantization": null, "latencyS": 0.733, "throughput": 89, "up": true }, { "provider": "AkashML", "input": 0.225, "output": 1.98, "blended": 0.66375, "cacheRead": 0.05, "context": 262144, "quantization": "fp8", "latencyS": 0.697, "throughput": 43, "up": true }, { "provider": "Parasail", "input": 0.24, "output": 2.2, "blended": 0.73, "cacheRead": 0.05, "context": 262144, "quantization": "fp8", "latencyS": 1.098, "throughput": 54, "up": true }, { "provider": "Reka AI", "input": 0.15, "output": 2.5, "blended": 0.7375, "cacheRead": 0.05, "context": 262144, "quantization": "fp8", "latencyS": 0.78, "throughput": 57, "up": true }, { "provider": "Mancer", "input": 0.175, "output": 2.5, "blended": 0.75625, "cacheRead": 0.175, "context": 262144, "quantization": "fp8", "latencyS": 0.984, "throughput": 56, "up": true }, { "provider": "ModelRun [by Modular]", "input": 0.2, "output": 2.5, "blended": 0.775, "cacheRead": 0.09, "context": 262144, "quantization": "fp4", "latencyS": 0.284, "throughput": 131, "up": true }, { "provider": "Phala", "input": 0.2, "output": 2.5, "blended": 0.775, "cacheRead": 0.05, "context": 1000000, "quantization": null, "latencyS": 1.198, "throughput": 76, "up": true }, { "provider": "Alibaba Cloud Int.", "input": 0.425, "output": 2.55, "blended": 0.95625, "cacheRead": 0.085, "context": 1000000, "quantization": null, "latencyS": 2.075, "throughput": 39, "up": true }, { "provider": "CoreWeave", "input": 0.4, "output": 3, "blended": 1.05, "cacheRead": 0.15, "context": 262144, "quantization": "fp8", "latencyS": 0.61, "throughput": 71, "up": true }, { "provider": "NovitaAI", "input": 0.42, "output": 3, "blended": 1.065, "cacheRead": 0.085, "context": 1000000, "quantization": null, "latencyS": 4.766, "throughput": 42, "up": true }, { "provider": "Nebius Token Factory", "input": 0.45, "output": 3, "blended": 1.0875, "cacheRead": 0.1, "context": 1000000, "quantization": "fp4", "latencyS": 1.122, "throughput": 123, "up": true }, { "provider": "Cerebras", "input": 0.99, "output": 1.49, "blended": 1.115, "cacheRead": 0.99, "context": 65536, "quantization": "fp16", "latencyS": 0.295, "throughput": 604, "up": true }, { "provider": "Cloudflare", "input": 0.45, "output": 3.2, "blended": 1.1375, "cacheRead": 0.05, "context": 262144, "quantization": null, "latencyS": 0.658, "throughput": 36, "up": true }, { "provider": "Venice", "input": 0.45, "output": 3.2, "blended": 1.1375, "cacheRead": null, "context": 262144, "quantization": "fp8", "latencyS": 1.15, "throughput": 62, "up": true } ] }, { "id": "openai/gpt-oss-120b", "name": "gpt-oss-120b", "author": "OpenAI", "url": "https://ernd.ai/inference-prices/gpt-oss-120b", "cheapest": "Prime Intellect", "endpoints": [ { "provider": "Prime Intellect", "input": 0.015, "output": 0.055, "blended": 0.025, "cacheRead": 0.013, "context": 131072, "quantization": "fp4", "latencyS": 0.84, "throughput": 73, "up": true }, { "provider": "NEAR AI", "input": 0.03, "output": 0.15, "blended": 0.06, "cacheRead": 0.015, "context": 131072, "quantization": "fp4", "latencyS": 0.639, "throughput": 49, "up": true }, { "provider": "Venice", "input": 0.03, "output": 0.15, "blended": 0.06, "cacheRead": 0.03, "context": 128000, "quantization": "int4", "latencyS": 0.832, "throughput": 40, "up": true }, { "provider": "CoreWeave", "input": 0.03, "output": 0.17, "blended": 0.065, "cacheRead": 0.03, "context": 131072, "quantization": "fp4", "latencyS": 0.269, "throughput": 91, "up": true }, { "provider": "DekaLLM", "input": 0.03, "output": 0.18, "blended": 0.0675, "cacheRead": 0.03, "context": 131072, "quantization": "bf16", "latencyS": 0.506, "throughput": 53, "up": true }, { "provider": "AkashML", "input": 0.03, "output": 0.187, "blended": 0.06925, "cacheRead": 0.03, "context": 131072, "quantization": "fp4", "latencyS": 0.39, "throughput": 103, "up": true }, { "provider": "DeepInfra (bf16)", "input": 0.037, "output": 0.17, "blended": 0.07025, "cacheRead": null, "context": 131072, "quantization": "bf16", "latencyS": 0.435, "throughput": 35, "up": true }, { "provider": "Crusoe", "input": 0.05, "output": 0.25, "blended": 0.1, "cacheRead": 0.05, "context": 131072, "quantization": "bf16", "latencyS": 0.238, "throughput": 217, "up": true }, { "provider": "Mancer", "input": 0.055, "output": 0.25, "blended": 0.10375, "cacheRead": 0.055, "context": 131072, "quantization": "fp8", "latencyS": 0.822, "throughput": 48, "up": false }, { "provider": "Sail Research", "input": 0.06, "output": 0.4, "blended": 0.145, "cacheRead": 0.03, "context": 131072, "quantization": "fp4", "latencyS": 0.624, "throughput": 129, "up": true }, { "provider": "DigitalOcean", "input": 0.06, "output": 0.42, "blended": 0.15, "cacheRead": 0.012, "context": 128000, "quantization": null, "latencyS": 0.69, "throughput": 51, "up": true }, { "provider": "Google Vertex (Global)", "input": 0.09, "output": 0.36, "blended": 0.1575, "cacheRead": null, "context": 131072, "quantization": null, "latencyS": 0.512, "throughput": 56, "up": false }, { "provider": "Baseten", "input": 0.1, "output": 0.5, "blended": 0.2, "cacheRead": 0.1, "context": 128072, "quantization": "fp4", "latencyS": 0.184, "throughput": 156, "up": true }, { "provider": "Baseten (US)", "input": 0.1, "output": 0.5, "blended": 0.2, "cacheRead": 0.1, "context": 128072, "quantization": "fp4", "latencyS": 0.168, "throughput": 165, "up": true }, { "provider": "Amazon Bedrock", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": null, "context": 131072, "quantization": null, "latencyS": 0.507, "throughput": 210, "up": true }, { "provider": "Amazon Bedrock (EU-WEST-1)", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": null, "context": 131072, "quantization": null, "latencyS": 0.484, "throughput": 152, "up": true }, { "provider": "Groq", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": 0.075, "context": 131072, "quantization": null, "latencyS": 0.188, "throughput": 342, "up": false }, { "provider": "Nebius Token Factory", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": 0.1, "context": 131072, "quantization": "fp4", "latencyS": 0.269, "throughput": 281, "up": true }, { "provider": "Phala", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": null, "context": 131072, "quantization": null, "latencyS": 0.933, "throughput": 145, "up": true }, { "provider": "SiliconFlow", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": 0.075, "context": 131072, "quantization": "fp8", "latencyS": 1.795, "throughput": 38, "up": false }, { "provider": "Together", "input": 0.15, "output": 0.6, "blended": 0.2625, "cacheRead": null, "context": 131072, "quantization": null, "latencyS": 0.342, "throughput": 91, "up": true }, { "provider": "Parasail", "input": 0.1, "output": 0.75, "blended": 0.2625, "cacheRead": 0.055, "context": 131072, "quantization": "fp4", "latencyS": 0.371, "throughput": 140, "up": true }, { "provider": "Parasail (Fast)", "input": 0.15, "output": 0.75, "blended": 0.3, "cacheRead": null, "context": 131072, "quantization": "fp4", "latencyS": 0.278, "throughput": 162, "up": true }, { "provider": "SambaNova", "input": 0.14, "output": 0.95, "blended": 0.3425, "cacheRead": null, "context": 131072, "quantization": null, "latencyS": 1.096, "throughput": 335, "up": true }, { "provider": "DeepInfra (fp8)", "input": 0.2, "output": 0.95, "blended": 0.3875, "cacheRead": null, "context": 131072, "quantization": "fp8", "latencyS": 1.423, "throughput": 59, "up": false }, { "provider": "Cerebras", "input": 0.35, "output": 0.75, "blended": 0.45, "cacheRead": 0.35, "context": 131072, "quantization": "fp16", "latencyS": 0.19, "throughput": 807, "up": true } ] } ] }