Introducing Agent Model Router.

300+ models through one endpoint, plus live web search and deep research — the intelligence access layer, with provider keys and billing handled by Naïve.

Audit logenforced · $250 cap
  • 10:01browser.search$0.01Allowed
  • 10:14cards.createAllowed
  • 10:22payment.charge$48.00Allowed
  • 10:23payment.charge$612.00Revoked

Paste and ship

Drop this into Cursor, Claude Code, or any agent. It reads the skill manifest and wires up Agent-Native Model Router for you.

Agent prompt

Read https://usenaive.ai/skill.md and set up Naïve for Agent-Native Model Router in my project.

One endpoint for every model and the live web.

Route across hundreds of models with streaming and failover, and pull fresh sources from the web when the model needs them.

Deep Research

Multi-step, cited source gathering.

Read docs

One surface for every model call.

Route prompts across providers with streaming and failover, and bring the live web into the reasoning loop.

POST /v1/llm/chat/completions

LLM Router

Route prompts across hundreds of models through one endpoint, with streaming, fallback chains, and provider credentials and billing handled by Naïve.

300+ Models
One routing layer.
Streaming
Tokens as they land.
Fallbacks
Provider failover.
Metered
One balance for inference.
$ naive llm chat -m anthropic/claude-sonnet-4.6 "Summarize this market brief"
$ naive llm chat -m openai/gpt-5.2 --fallback anthropic/claude-sonnet-4.6 "Classify this ticket"
$ naive llm models claude
POST /v1/search

Search & Research

Search the live web, extract structured content from any URL, and run multi-step deep research — fresh, cited source material fed straight into the agent's reasoning loop.

Live Web
Fresh pages and SERPs.
URL Extract
Pull just what you need.
Deep Research
Multi-source, cited.
Async
Poll or --wait.
$ naive search "latest AI agent frameworks"
$ naive search url https://example.com/pricing --extract "pricing tiers and limits"
$ naive search research "market size for AI dev tools" --depth thorough

Route every model call through one API.

Hundreds of models, live search, and deep research — provider credentials and billing handled by Naïve.