Models

30+ models from 6 providers. Starting at 0.01 credits per call. We route to the most cost-effective provider — you just pick the model.

Cost-effective routing

We route each model to the most cost-effective provider automatically. You get the best price without managing provider accounts.

One line to switch

Switch models with one parameter: model: "claude-sonnet-4.6". Override provider: model: "llama-3.1-8b@groq"

Pay only for what you use

Credits based on actual usage. We estimate before each call and automatically refund unused credits after.

How credits work

Credits are the universal currency across all models. A short message on a budget model costs as little as 0.01 credits. Longer conversations on premium models cost more.

The table below shows credit ranges (short prompt to long conversation). We estimate before each call and automatically refund unused credits after.

Model Provider Credits (typical) Context Tier
claude-sonnet-4.6 Anthropic 0.5 - 8 200K Premium
claude-opus-4.6 Anthropic 2 - 40 200K Premium
gpt-4.1 OpenAI 0.5 - 5 1M Premium
gpt-4o OpenAI 0.5 - 6 128K Premium
o3 OpenAI 0.5 - 5 200K Premium
o4-mini OpenAI 0.1 - 3 200K Premium
o3-mini OpenAI 0.1 - 3 200K Premium
claude-sonnet-4 Anthropic 0.5 - 8 200K Premium
grok-3 xAI 0.5 - 8 131K Premium
claude-haiku-4.5 Anthropic 0.1 - 3 200K Standard
gpt-4.1-mini OpenAI 0.05 - 2 1M Standard
gemini-2.5-flash Google 0.05 - 2 1M Standard
gpt-4o-mini OpenAI 0.01 - 0.5 128K Standard
llama-3.3-70b Groq 0.01 - 1 128K Standard
deepseek-r1-32b Workers AI 0.1 - 3 128K Standard
gpt-oss-120b Groq 0.01 - 0.5 128K Standard
grok-4.1-fast xAI 0.01 - 0.5 128K Standard
grok-3-mini xAI 0.01 - 0.5 131K Standard
llama-3.1-8b Groq 0.01 - 0.1 131K Budget
gemini-2.5-flash-lite Google 0.01 - 0.5 1M Budget
gpt-4.1-nano OpenAI 0.01 - 0.5 1M Budget
claude-3-haiku Anthropic 0.01 - 1 200K Budget
gemini-2.0-flash Google 0.01 - 0.5 1M Budget
gpt-oss-20b Groq 0.01 - 0.3 128K Budget
llama-3.2-3b Workers AI 0.01 - 0.5 128K Budget
llama-3.2-1b Workers AI 0.01 - 0.5 128K Budget
mistral-7b Workers AI 0.01 - 0.5 32K Budget
phi-2 Workers AI 0.01 - 0.5 2K Budget

Credit estimates are for a typical request (~500 input + 1024 output tokens). Actual credits depend on your usage. Short prompts cost less; long conversations cost more. Unused credits are refunded after each call.

Buy credits · API docs