ernd guides
Guides to AI API costs
How OpenAI and Anthropic bill for their APIs and how to keep that bill down. There are guides to getting your rate limits raised, too. Every fact links to the vendor's own page, so you can check it there.
Guides
13 pages
-
How to cut LLM API costs
How to cut LLM API costs on OpenAI and Anthropic. Start with a smaller model and prompt caching, then move work that can wait to batch and set spend limits.
-
How OpenAI and Anthropic API billing works
How OpenAI and Anthropic API billing works, from prepaid credits and auto-reload to usage tiers, spend limits, invoices and tax. Checked October 27, 2026.
-
How to track AI API spend
How to track AI API spend on OpenAI and Anthropic with usage dashboards, cost APIs, per-project keys and spend alerts, then reconcile it with receipts.
-
AI credits programs for startups, compared
AI credits programs for startups from OpenAI, Anthropic, AWS, Google Cloud and Microsoft, compared on amounts, who qualifies, expiry and the models covered.
-
How to get higher rate limits from AI providers
How to get higher rate limits from AI providers: how OpenAI, Anthropic, Gemini, Azure OpenAI and Amazon Bedrock set tiers, and how to ask for more.
-
Business credit cards for AI spend, compared
Business credit cards for AI spend, compared: what Amex, Ramp, Mercury and the ernd card give a US company on its AI bills, from ChatGPT credits to cash back.
-
How to raise your Amazon Bedrock quotas
How to raise your Amazon Bedrock quotas per model and Region, how requests differ on bedrock-runtime and bedrock-mantle, and why AWS might turn one down.
-
How to raise your Anthropic rate limits
How to raise your Anthropic rate limits with Request tier increase in the Claude Console, how tiers and spend caps work, and what to do when you hit one.
-
How to raise your Azure OpenAI quota
How to raise your Azure OpenAI quota: how subscription tiers move up, how to use the quota request form, and when provisioned throughput fits better.
-
How to raise your Gemini API rate limits
How to raise your Gemini API rate limits: how usage tiers work, what counts as spend for Tier 2 and Tier 3, and which form to use when you need more.
-
How to raise your Mistral rate limits
How to raise your Mistral rate limits: where your API limits are listed, what to send support when you ask for more, and how spending limits work.
-
How to raise your OpenAI rate limits
How to raise your OpenAI rate limits by buying credits to reach the next usage tier, and what to do once Grow, the top self-serve tier, isn't enough.
-
How to raise your xAI rate limits
How to raise your xAI rate limits: tiers follow cumulative spend since January 1, 2026. What counts as spend, how RPS is set, and when to ask sales.
Pricing and tools
4 pages
-
OpenAI API pricing per million tokens
OpenAI API pricing per 1M tokens for GPT-6 and GPT-5.6 models, with cached input, Batch, Flex and Fast mode rates, as listed on October 27, 2026.
-
Anthropic API pricing per million tokens
Anthropic API pricing per million tokens for Claude Opus, Sonnet, Haiku and Fable models, with cache, Batch and fast mode rates, as listed on October 27, 2026.
-
AI API cost calculator for OpenAI and Anthropic
Estimate a month of OpenAI or Anthropic API spend from token counts, at list prices checked October 27, 2026. Examples only, not a quote.
-
AI API rate limits
What 13 AI vendors limit, what moves you up a tier and how to ask for more. Every number links to the vendor's own page.
More on billing and pricing is on the ernd blog.
The ernd card is a credit card for US companies, with partners across the AI stack built in. Request your card. ernd is run by Provenance Research Inc.
The ernd card
The ernd card is a credit card for US companies, with partners across the AI stack built in.
Request your card