Skip to content

Overview — unlimited LLM API at a flat monthly price

CheapestInference is an AI inference platform that gives you access to open-source models through a single API with flat monthly pricing. No per-token charges.

Six models are served across two pools: the Frontier Pool (Kimi K2.7, Kimi K2.6, GLM 5.2, MiniMax M3) and the Core Pool (DeepSeek V4 Flash, MiMo v2.5). A subscription is per-pool — during your reserved time blocks you get unlimited usage of every model in that pool. See Models for the full catalog.

https://api.cheapestinference.com/v1

CheapestInference serves your requests against the open-source models in your pool.

  1. You subscribe to a pool by reserving one or more daily time blocks
  2. You create API keys from the dashboard
  3. You use those keys with the OpenAI or Anthropic SDK — just change the base URL
  4. The platform validates your key and serves the request against the requested model

Your API key works exactly like an OpenAI or Anthropic key. All routing and spend tracking is handled automatically.

EndpointDescription
POST /v1/chat/completionsOpenAI-compatible chat (all models)
POST /anthropic/v1/messagesAnthropic-compatible messages
GET /v1/modelsList available models
GET /v1/models/:model_idGet specific model details
GET /v1/usageCheck key usage and status

The response format matches the official OpenAI and Anthropic APIs exactly.

Each key handles a limited number of simultaneous requests during your reserved blocks. To run more in parallel — or to isolate clients — create additional keys (one per seat). Keys are independent, so one busy key never affects the others.

MethodHow
CardVisa, Mastercard, etc. via Stripe
USDCDirect transfer on Base L2 (MetaMask, Coinbase Wallet)

Card subscriptions renew automatically until you cancel. USDC subscriptions are one-time (30 days) — you renew manually when ready.

Requests without an API key receive a 402 Payment Required response with a product catalog. AI agents can subscribe autonomously using the x402 protocol with USDC on Base L2 — no human setup needed.

No custom SDK required. Use the official OpenAI or Anthropic SDK in any language:

  • Python: openai, anthropic
  • Node.js: openai, @anthropic-ai/sdk
  • Any OpenAI-compatible client (Go, Rust, Java, etc.)