Unlimited FREE AI coding. Connect Claude Code, Codex, Cursor, Cline, Copilot, Antigravity to FREE Claude/GPT/Gemini via 40+ providers. Auto-fallback, RTK -40% tokens, never hit limits.

Never stop coding. Free AI gateway: one endpoint, 160+ providers (50+ free), connect Claude Code, Codex, Cursor, Cline & Copilot to FREE Claude/GPT/Gemini. RTK+Caveman stacked compression saves 15-95% tokens, smart auto-fallback, MCP/A2A, multimodal APIs, Desktop/PWA.

Project-agnostic dual-memory MCP CLI for Claude Code, Cursor, and OpenCode (Qdrant tuned hybrid retrieval + structural memory hooks)

Smart AI proxy router that connects Claude Code, Cursor, Codex, Copilot, and 50+ coding tools to 40+ AI providers with 100+ models. Saves 20-40% tokens via RTK compression, auto-fallback from subscription to cheap to free providers, multi-account round-robin, and a local dashboard.

Real-time dashboard to track both per-project and total token usage for Claude Code, providing recommendations for token optimization and improving response quality.

Stop hitting Claude Pro's weekly limit by Wednesday. Delegate routine & complex tasks from Claude Opus 4.7 to DeepSeek – slash commands included.

Explore codebases with web LLMs — no API keys, just manual copy-paste (no browser automation). Ideal for pre-dev research: search, analyze, and generate docs while saving massive tokens. For heavy editing, use Claude Code, Codex or other similar tools instead. Zero API cost, fully ToS-compliant.

ContextSlimAI - A powerful tool to optimize and reduce LLM context window size, saving tokens and improving AI agent efficiency.

Reject the Token-Tax on noise. A sanctuary for Smart Shrimps to harvest pure logic. No UI, 404 by default, Signal by design

Windows AI workbench token-saver patch kit: audit, evidence manifests, preflight gates, and AgentWorkOS assets.

MOS — AI optimization layer for Claude Code. Easy install: npx mos-claude | Auto-scores task complexity (0-140), matches model/thinking/subagents to needs. Cut token costs 30-75%.