|
|
Multi-agent pipeline where LLMs work as a team: gatekeeper → planner → workers → verifier. Weak models get stronger through strict JSON contracts and automatic answer correction — retries that know why they retry. |
|
|
Local OpenAI-compatible proxy for Qwen Chat: any model ID passes through as-is, multi-account rotation, files, image & video generation. Plugs into Open WebUI, LiteLLM, coding agents. |
|
|
HTTP bridge that exposes the Cline coding agent as an OpenAI-compatible API: SSE streaming, file attachments, optional agent tools — zero runtime dependencies. |
|
|
Web dashboard for token usage collected from local AI tools (Hermes, OpenCode). Reads only your own SQLite — if tokens flow, they get counted. |
Popular repositories Loading
-
TokenEater
TokenEater PublicExperimental multi-agent LLM pipeline on free OpenRouter models with automatic answer verification
Python 2
-
FreeQwenApi
FreeQwenApi PublicForked from ForgetMeAI/FreeQwenApi
Local OpenAI-compatible proxy for Qwen Chat: any model ID passes through, multi-account rotation, files, image & video generation. Works with Open WebUI / LiteLLM / agents
JavaScript 1
-
-
cline-api
cline-api PublicOpenAI-compatible HTTP bridge for Cline with SSE streaming, file attachments, and optional agent tools
JavaScript
-
llm-token-dashboard
llm-token-dashboard PublicWeb dashboard for LLM token usage from local AI tools (Hermes / OpenCode) - SQLite + Flask, fully offline
HTML
If the problem persists, check the GitHub status page or contact support.