
claude-model-router
A Claude Code skill that picks the right Claude model for each piece of work.
Install with your AI
Paste into Claude Code, Cursor, or any agent — it reads the repo and wires the tool into your project.
Install and set up claude-model-router (claude-skill project) into my current project. Found on https://claudeers.com/claude-model-router Repo: https://github.com/Rajvardhman05/claude-model-router Homepage/docs: — Detected install method: claude-skill → # copy this skill into .claude/skills/claude-model-router/ Category: other. Platforms: cli. Read the repo's README for exact setup and env vars, then install it and wire it into my project. Claudeers Health Verdict: unknown; community-verified: false. Confirm the source before running anything.
# copy the skill dir into your project: # .claude/skills/claude-model-router/ (or ~/.claude/skills/claude-model-router/ for all projects)
git clone https://github.com/Rajvardhman05/claude-model-router
// compatibility
| Platforms | cli |
|---|---|
| Operating systems | — |
| AI compatibility | claude |
| License | MIT |
| Pricing | open-source |
| Language | — |
claude-model-router
A Claude Code skill that picks the right Claude model for each piece of work, so you stop running everything on one.
Anthropic published a guide on which model to use for which task. Nobody is going to consult a table before every prompt. This skill hands the table to Claude: the main session stays on your model, and whenever it delegates to a subagent it picks the cheapest tier that will finish the job. Haiku reads the logs, Sonnet makes the scoped edit, Fable plans the migration.
What it does
- Routes subagents by task type (haiku / sonnet / opus / fable).
- Never switches your main model mid-session, because that throws away the prompt cache.
- Escalates in the right order when a result falls short: verify, then effort, then model.
- Flags a clear mismatch between your session's model and the task, in one line.
Install
git clone https://github.com/Rajvardhman05/claude-model-router ~/.claude/skills/claude-model-router
Then add this to ~/.claude/CLAUDE.md so the routing fires every time, not only when the skill
happens to load:
## Model Routing
Before spawning any subagent, load the `claude-model-router` skill and set the subagent's model from
its routing table. Never switch the main session's model mid-task.
Source
Rules are condensed from Anthropic's "Which Claude model should you build with?" (https://claude.dev/blog/which-claude-model-should-you-build-with/, November 2026). Model tiers and prices change; re-check the guide when a new generation ships.
License
MIT
// faq
What is claude-model-router?
A Claude Code skill that picks the right Claude model for each piece of work.. It is open-source on GitHub.
Is claude-model-router free to use?
claude-model-router is open-source under the MIT license, so it is free to use.
What category does claude-model-router belong to?
claude-model-router is listed under other in the Claudeers registry of Claude-compatible tools.
// embed badge
[](https://claudeers.com/claude-model-router)
// retro hit counter
[](https://claudeers.com/claude-model-router)
// reviews
// guestbook
// related in Other
Anti-AI-slop design skill for Claude Code, Cursor, and Codex.
Huashu Design · HTML-native design skill for Claude Code · Claude Code 里 HTML 原生的设计 skill · 高保真原型 / 幻灯片 / 动画 + 20 设计哲学 + 5 维评审 + MP4 导出 · Agent-agnostic
Open source Ghostty-based macOS terminal with vertical tabs and notifications for AI coding agents. Built for multitasking, organization, and programmability.