LLM

Give your agent 300+ models through one API

Route agent reasoning through a single API with model fallbacks, usage metering, and the same bearer token as every other primitive.

Read the launch post →
Run in your agent

Paste and ship

Drop this prompt into Cursor, Claude Code, or any coding agent. It reads the skill manifest and wires up LLM for you.

Agent prompt

Read https://usenaive.ai/skill.md and use the Naïve LLM primitive in my project. Start with: naive llm chat -m anthropic/claude-sonnet-4.6 "..."

Drop into any agent

In action

Chat completions across 300+ models through one router.

  • 300+ modelsOpenAI, Anthropic, Google, open weights.
  • One routerSwap models without new integrations.
  • Metered usageDraws from your credit balance.
  • Tool-readyDrop into any agent runtime.
Exact code

One call to llm

Same primitive from the CLI, SDK, or REST — one bearer token, one credit balance.

CLI

LLM

Route agent reasoning through a single API with model fallbacks, usage metering, and the same bearer token as every other primitive.

300+ models
OpenAI, Anthropic, Google, open weights.
One router
Swap models without new integrations.
Metered usage
Draws from your credit balance.
Tool-ready
Drop into any agent runtime.
# Run with your API key exported
$ naive llm chat -m anthropic/claude-sonnet-4.6 "..."

Related primitives

Start building with LLM

Get your API key and add llm to any agent in minutes — no infrastructure to run.