is now an ernd partner: 10% back on inference On this page
Cheapest and fastest
- Cheapest
- Wafer (US)ernd partner
- $2.61 per 1M tokens blended
- $0.276 input, $9.60 output
- Fastest
- Parasail (Fast)
- 106 tokens per second
- $9.00 blended, 0.55 s to first token
Every provider
| Provider | Input /1M | Output /1M | Blended 3:1 | Cache read /1M | Context | Quantization | Latency | Throughput | With ernd card |
|---|---|---|---|---|---|---|---|---|---|
| Wafer (US) ernd partner | $0.276 | $9.60 | $2.61 | $0.13 | 1.05M | MXFP4 | 0.76 s | 74 tok/s | $2.35after 10% cash back |
| Relace | $0.166 | $10.14 | $2.66 | $0.16 | 1.05M | FP4 | 1.35 s | 65 tok/s | – |
| Makora | $0.36 | $12.00 | $3.27 | $0.16 | 1.05M | FP4 | 1.30 s | 53 tok/s | – |
| Sail Research | $0.276 | $13.26 | $3.52 | $0.138 | 1.05M | FP4 | 1.20 s | 76 tok/s | – |
| Pareto Alpha (US) | $1.00 | $11.25 | $3.56 | $0.225 | 1.05M | Not listed | 6.91 s | 23 tok/s | – |
| Sail Research (US) | $0.281 | $13.50 | $3.59 | $0.14 | 1.05M | FP4 | 1.03 s | 76 tok/s | – |
| Chutes | $1.85 | $9.50 | $3.76 | $0.185 | 1.05M | MXFP4 | 3.13 s | 18 tok/s | – |
| Decart | $1.95 | $9.75 | $3.90 | $0.195 | 1.05M | MXFP4 | 1.81 s | 66 tok/s | – |
| inference.net | $0.234 | $15.00 | $3.93 | $0.195 | 1.05M | FP4 | 0.88 s | 53 tok/s | – |
| inference.net (US) | $0.234 | $15.00 | $3.93 | $0.195 | 1.05M | FP4 | 0.90 s | 57 tok/s | – |
| Wafer (Fast) ernd partner | $0.155 | $16.00 | $4.12 | $0.154 | 1.05M | MXFP4 | 0.88 s | 74 tok/s | $3.70after 10% cash back |
| Wafer ernd partner | $0.192 | $16.00 | $4.14 | $0.182 | 1.05M | MXFP4 | 0.86 s | 80 tok/s | $3.73after 10% cash back |
| Morph | $2.17 | $12.00 | $4.63 | $0.534 | 1.05M | FP8 | 1.52 s | 64 tok/s | – |
| inference.net (Fast) | $0.45 | $18.00 | $4.84 | $0.315 | 1.05M | FP4 | 0.89 s | 75 tok/s | – |
| inference.net (US, Fast) | $0.45 | $18.00 | $4.84 | $0.315 | 1.05M | FP4 | 0.93 s | 74 tok/s | – |
| Phala | $2.55 | $12.75 | $5.10 | $0.255 | 1.05M | Not listed | 4.26 s | 26 tok/s | – |
| DigitalOcean | $2.55 | $12.95 | $5.15 | $0.255 | 1.05M | Not listed | 1.66 s | 29 tok/s | – |
| io.net | $2.30 | $14.00 | $5.22 | $0.23 | 1.05M | FP4 | 2.44 s | 39 tok/s | – |
| DeepInfra | $2.85 | $14.25 | $5.70 | $0.285 | 1.05M | MXFP4 | 1.16 s | 38 tok/s | – |
| Amazon Bedrock (US-EAST-2) | $3.00 | $15.00 | $6.00 | $0.30 | 1.05M | Not listed | 4.05 s | 27 tok/s | – |
| Baseten | $3.00 | $15.00 | $6.00 | $0.30 | 1.05M | FP8 | 1.53 s | 63 tok/s | – |
| Fireworks | $3.00 | $15.00 | $6.00 | $0.30 | 1.05M | Not listed | 1.79 s | 48 tok/s | – |
| Modal | $3.00 | $15.00 | $6.00 | $0.30 | 1.05M | MXFP4 | 0.72 s | 82 tok/s | – |
| Moonshot AI | $3.00 | $15.00 | $6.00 | $0.30 | 1.05M | Not listed | 5.47 s | 29 tok/s | – |
| Parasail | $3.00 | $15.00 | $6.00 | $0.30 | 1.05M | FP4 | 1.46 s | 47 tok/s | – |
| Together | $3.00 | $15.00 | $6.00 | $0.30 | 1.05M | Not listed | 2.73 s | 33 tok/s | – |
| Alibaba Cloud Int. | $3.45 | $17.25 | $6.90 | $0.345 | 1.05M | Not listed | 2.17 s | 30 tok/s | – |
| Fireworks (Fast) | $4.50 | $22.50 | $9.00 | $0.45 | 1.05M | Not listed | 1.34 s | 76 tok/s | – |
| Fireworks (US) | $4.50 | $22.50 | $9.00 | $0.45 | 1.05M | Not listed | 1.13 s | 63 tok/s | – |
| Parasail (Fast) | $4.50 | $22.50 | $9.00 | $0.45 | 1M | FP4 | 0.55 s | 106 tok/s | – |
Blended is three parts input price to one part output, per million tokens. Latency is time to first token and throughput is output tokens per second, both averaged from OpenRouter's daily figures for the last 7 days. Cheapest and fastest only count listings that were up when checked.
Wafer is an ernd partner: 10% back on Wafer inference with the ernd card. With ernd card is the blended price less 10% cash back on Wafer charges. It comes back on the card bill and doesn't change Wafer's list price. It applies when you pay Wafer directly, not through OpenRouter credits.
Questions
What is the cheapest Kimi K3 API provider?
Wafer (US), at $2.61 per million tokens blended ($0.276 input, $9.60 output), in OpenRouter's listings on November 7, 2026. If your workload is mostly output, Chutes has the lowest output price at $9.50.
How much does Kimi K3 cost per million tokens?
Input runs from $0.155 to $4.50 per million tokens across 30 listings, and output from $9.50 to $22.50. The middle listing comes to $4.97 blended.
Which Kimi K3 provider is fastest?
Parasail (Fast) led on both counts over the last 7 days, at about 106 tokens per second and 0.55 s to the first token.
How do I pay for Kimi K3 inference?
Buy OpenRouter credits and it sends each request to one of these providers, or open an account with a provider and pay it directly. The ernd card is a credit card for US companies. It earns 4% back on AI vendor spend and 1.5% back on everything else. Wafer charges on the card earn 10% back.
More models
- DeepSeek V4.1 Flash 1017 from $0.0813 blended
- DeepSeek V4.1 Flash from $0.0887 blended
- GLM 5.3 Flash from $0.0459 blended
- Hy4 preview from $1.13 blended
- MiniMax M3.1 Flash from $0.131 blended
- DeepSeek V4 Flash 0731 from $0.0667 blended
- GLM 5.3 from $0.402 blended
- Kimi K3.1 Pro from $2.27 blended
- Nemotron 3 Ultra from $0.74 blended
- Hy3 from $0.144 blended
- Nemotron 3.5 Super VL from $0.133 blended
- MiMo-V2.6-Flash from $0.14 blended
- DeepSeek V4 Flash 0423 from $0.0544 blended
- Nemotron 3.5 Lightning from $0.0488 blended
- MiMo-V2.6-Pro from $0.486 blended
- GLM 5.2 from $0.29 blended
- DeepSeek V4 Pro 0813 from $0.871 blended
- Qwen3.8 27B from $0.342 blended
- gpt-oss-120b from $0.025 blended
Source
Checked . Providers change prices often, so OpenRouter's page is the final word. ernd is run by Provenance Research Inc.