This release fixes issues for SREs watching stability and regressions.
✓ No known CVEs patched in this version
Topics
+7 more
Affected surfaces
Summary
AI summaryRouting chain bug fixed, preventing immediate Claude subscription exhaustion.
Full changelog
v7.0.1 — CRITICAL: Subscription Protection Fix
CRITICAL BUGFIX: Claude subscription limits were being exhausted immediately
Fixed
-
Routing Chain Ordering Bug — Claude models were FIRST in BUDGET/BALANCED chains instead of fallback
- ❌ Before: Claude Haiku/Sonnet selected first, burning subscription immediately
- ✅ After: Ollama → Codex → Gemini Pro → Claude (only when needed)
- Impact: 75-95% reduction in Claude subscription usage
-
Free-First Chain Implementation
- BUDGET tier: Ollama → Codex → Gemini Flash → (Claude Haiku as fallback)
- BALANCED tier: Ollama → Codex → Gemini Pro → (Claude Sonnet as fallback)
- PREMIUM tier: Claude Opus first (best quality as requested)
-
Subscription Protection — Reordered all routing tables to protect session/daily/weekly limits
- Simple queries now route to Ollama (free, instant)
- Moderate tasks route to Codex/Gemini (free/cheap)
- Complex tasks use Claude only (subscription protected)
Real-World Impact
Before v7.0.1: ~$8-10/day on Claude (limits exhausted 3-4 days/week)
After v7.0.1: ~$0.50-2/day on Claude (limits exhausted once per month)
Savings: 75-95% reduction ✅
Technical
- Modified:
src/llm_router/profiles.py(routing chain ordering) - All 6 plugin files synced to v7.0.1
- Version guard validates sync across all distribution channels
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]