This release adds 2 notable features for engineering teams evaluating rollout.
✓ No known CVEs patched in this version
Topics
+7 more
Affected surfaces
Summary
AI summaryFixed silent ignoring of routing directives and redefined LLM_ROUTER_CLAUDE_SUBSCRIPTION behavior.
Full changelog
What's fixed
The LLM_ROUTER_CLAUDE_SUBSCRIPTION=true mode was emitting /model claude-haiku-4-5-20251001 directives that Claude Code's model cannot execute — slash commands only work when typed by a human in an interactive terminal. Every CC-MODE simple/complex routing directive was silently ignored.
Fix: always route via MCP tools
All tasks now route via MCP tools (free-first chain: Ollama → Codex → cheap API), regardless of subscription mode. This actually works in all contexts — interactive, claude -p, Docker, agents.
Forward compatibility: LLM_ROUTER_MODEL_SWITCH=true
The /model path is preserved but disabled by default. When Claude Code gains the ability to execute slash commands from hook context, re-enable it:
export LLM_ROUTER_MODEL_SWITCH=true
LLM_ROUTER_CLAUDE_SUBSCRIPTION=true now means
- Inline OAuth refresh (keeps subscription usage data fresh)
- Session-end delta reporting (shows how much subscription quota was used)
- Does not change routing behaviour
Upgrade
pip install --upgrade claude-code-llm-router
llm-router install --force
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
Share this release
About ypollak2/llm-router
Subscription-aware LLM router for Claude Code. Routes tasks to 20+ providers (OpenAI, Gemini, Groq, Ollama, Codex) based on complexity classification, Claude subscription pressure, and cost. Free tasks stay on Claude subscription; expensive tasks fall back to the cheapest capable model. Includes 30 MCP tools, 6 auto-routing hooks, semantic dedup cache, prompt caching, daily spend cap, and a live web dashboard.
Related context
Related tools
Beta — feedback welcome: [email protected]