# xllm-go/bypass

集成了openai-api、coze、deepseek、cursor、windsurf、qodo、blackbox、you、grok、bing  绘画 多款AI的聊天逆向接口适配到 OpenAI API 标准接口服务端。

Repository: https://github.com/xllm-go/bypass
Canonical: https://ross.abutalabs.com/products/xllm-go-bypass
Homepage: https://xllm-go.github.io/bypass
Language: Go
License: GPL-3.0
License Family: copyleft
Topics: ai, claude, gpt, coze, you, bing, deepseek, cursor, grok, windsurf, qodo
Last push: 2026-06-24T22:52:43+00:00

## Health v2 (maintenance only)
Score: 62/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 89, release rhythm 16, longevity 82
- inputs: {"age_days": 1161, "days_push": 70, "days_rel": 573, "gap_med": 150, "n_releases_24m": 2}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1239, forks 233 (observed 2026-08-28T04:04:06.037757+00:00)

## What it is
A Go server that reverse-engineers the chat interfaces of multiple AI providers (Coze, DeepSeek, Cursor, Windsurf, Grok, Bing Copilot, You, Qodo, Blackbox, HuggingFace image generation) and exposes them through a standard OpenAI-compatible v1 API. It supports streaming output and multi-turn conversation, allowing any OpenAI API client to use these backends.

## Use cases
- use free AI chat backends through an OpenAI-compatible API
- expose Bing Copilot or Grok as an OpenAI v1 endpoint
- connect Cursor or Windsurf models to my own chat client
- self-host an OpenAI API proxy for DeepSeek and Coze
- generate images via HuggingFace through an OpenAI-style API
- run a streaming multi-turn chat server compatible with ChatGPT clients

## When to choose
- you want a single OpenAI-compatible endpoint aggregating many free/reverse-engineered AI chat services
- you need streaming and multi-turn conversation support with existing OpenAI SDK clients
- you want a self-hostable Go service with Docker deployment

## When to avoid
- you need a fully legal, officially supported API with provider SLAs
- you require guaranteed stability, since reverse-engineered endpoints break frequently
- you plan commercial use, which the project explicitly prohibits

## Facets
- artifact type: service
- maturity: active
- function: http-server, api-framework, llm-inference, chatbot, reverse-engineering, web-scraping
- domain: large-language-models, artificial-intelligence, apis, self-hosted, chatbots
- platform: windows, self-hosted, go, cli
- tags: openai-api-compat, reverse-proxy, chatgpt-adapter, api-adapter, free-ai-apis, streaming, image-generation, linux, macos, docker

## Member repositories
- xllm-go/bypass (main) score 62

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:06.037757+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T05:10:00.224189+00:00, confidence not recorded.
  - readme: https://github.com/xllm-go/bypass (fetched 2026-08-28T04:04:06.037757+00:00, sha 6c7fda4e628a)
  - homepage: https://xllm-go.github.io/bypass (fetched 2026-08-29T12:20:41.922806+00:00, sha 880347349ae3)
- Data as of 2026-08-30T08:39:29.467469+00:00.