InferSpeed Pay-per-token model API with OpenAI- and Anthropic-compatible endpoints. OpenAI base URL https://api.inferspeed.ai/v1, Anthropic base URL https://api.inferspeed.ai, header Authorization: Bearer $INFERSPEED_API_KEY. Models on sale and live prices: https://inferspeed.ai/model-plaza (JSON: /api/v1/public/pricing). This site needs JavaScript. Plain-text versions for people and agents:

/developers.md - API settings and setup for Codex CLI, Claude Code, OpenCode and SDKs

/agents.md - instructions for an agent configuring itself

/llms.txt - site index

/skill.md - agent skill: settings, prices, balance, errors, key handling

/setup.sh - installer for Claude Code, Codex CLI, OpenCode and OpenClaw

/api/v1/public/pricing - prices as JSON

Run any agent on cheaper models

InferSpeed is a pay-per-token API for AI language models: a service you sign up for, then call with an API key through OpenAI- and Anthropic-compatible endpoints, billed per token from a prepaid balance.

Setup guides

Docs

Quickstart: base URLs, keys and tested configs
Models and prices
llms.txt: plain-text index for language models (/llms-full.txt in one file)
developers.md: the API page as Markdown

Models