ModelFlux: OpenAI-compatible model traffic router with health-aware key-pool scheduling and failover
A serverless, OpenAI-compatible LLM API gateway for Cloudflare Workers — a new-api / one-api alternative that runs on the free tier (Workers + D1, no server).
Self-hosted LLM gateway — from LAN to the public internet (tunnel-friendly): multi-upstream key pool, downstream tokens, quota accounting, MCP & A2A routing, E2E envelope encryption. 自托管大模型中转站，局域网与公网两用，兼容 OpenAI/Anthropic。
🏊 Hard fork of new-api with integrated Key Pool Management — first-class menu for upstream LLM key lifecycle: accounts, automation recipes, billing, alerts.
Active health guardian for LLM API key pools — probe every key, quarantine the bad ones, re-admit on recovery. A tiny sidecar beside your existing gateway (litellm/one-api/gpt-load), not another gateway.
High-resilience multi-LLM orchestration engine featuring AES-256 encrypted key pool leasing, multi-provider cascading failovers, ID3-stripped TTS synthesis, and real-time telemetry.
DeepSeek Harness plugin: pool several OpenCode Go subscriptions behind one provider route, serve each request with the healthiest key, and fail over when a quota window is spent.
Chii (agy-router): Multi-Key Pool Router & OpenAI-Compatible Local Adapter for Google Antigravity Interactions API
Stdlib-only Python CLI coding agent — 12 AI providers, agentic tool loop, SQLite sessions, MCP support
DeepSeek Harness plugin: OpenCode Go key pool with session-pinned rotation
把多个 Tavily API key 汇成单一入口的轻量网关：MCP + Tavily 兼容 REST、按剩余额度选上游、401/429/432/433 自动换 key、响应缓存与健康检查（只含配置与脚本，不含凭据）
OpenCode Go API key pool for DeepSeek Harness: sticky keys, quota failover, official 5h/week/month usage in Settings
Self-hosted Exa API gateway with key pooling, failover, encrypted storage, and admin console.
To associate your repository with the key-pool topic, visit your repo's landing page and select "manage topics."