# james-6-23/codex2api

Codex2API 是一个基于 Go + Gin + React/Vite 的 Codex 反向代理与管理后台项目

Repository: https://github.com/james-6-23/codex2api
Canonical: https://ross.abutalabs.com/products/codex2api
Homepage: https://codex2api-latest-vu8j.onrender.com
Language: Go
License Family: other
Topics: 2api, codex
Last push: 2026-08-26T19:43:24+00:00

## Health v2 (maintenance only)
Score: 77/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 99, release rhythm 87, longevity 11
- inputs: {"age_days": 163, "days_push": 7, "days_rel": 7, "gap_med": 0, "n_releases_24m": 128}
- flags: young, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1983, forks 350 (observed 2026-08-28T04:06:01.820755+00:00)

## What it is
Codex2API is a self-hosted Go/Gin reverse proxy with an embedded React admin dashboard that turns a pool of Codex accounts into an OpenAI- and Anthropic-compatible API gateway. It exposes chat completions, responses, messages, images, and models endpoints while handling account scheduling, health scoring, dynamic concurrency, rate-limit recovery, and usage tracking.

## Use cases
- expose an OpenAI-compatible API from a Codex account pool
- point Codex CLI or Claude Code at one self-hosted base URL
- rotate multiple Codex accounts with quota-aware load balancing
- add usage tracking and health monitoring to a Codex-to-API bridge
- self-host an Anthropic Messages-compatible gateway backed by Codex
- manage API keys, proxies, and prompt filtering from an admin console
- run a lightweight SQLite single-container gateway for Codex access

## When to choose
- you have multiple Codex accounts and need one observable, schedulable OpenAI/Anthropic-compatible endpoint
- you want built-in account health tiers, cooldown recovery, round-robin or remaining-quota scheduling
- you need an embedded admin dashboard for accounts, keys, proxies, and usage analytics
- you want flexible deployment from a single SQLite container up to a PostgreSQL + Redis stack

## When to avoid
- you need a thin, minimal proxy with no account-pool management or dashboard
- you require a permissively licensed dependency since the repository has no license
- you depend on officially supported OpenAI API access rather than unofficial Codex account forwarding
- you need actual model inference or fine-tuning, since this only forwards requests to Codex

## Facets
- artifact type: service
- maturity: active
- function: api-gateway, proxy, http-server, auth, rate-limiting, monitoring, self-hosted
- domain: large-language-models, artificial-intelligence, apis, developer-tools, self-hosted, backend
- platform: go, self-hosted, cross-platform
- tags: openai-compatible, anthropic-compatible, codex, reverse-proxy, account-pool, llm-gateway, admin-dashboard, api-key-management, token-refresh, usage-tracking, health-scoring, gin, react-dashboard, sqlite, postgresql, redis, docker, linux, web-server

## Member repositories
- james-6-23/codex2api (main) score 77

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:01.820755+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 2026-08-30T03:03:30.329051+00:00, confidence not recorded.
  - readme: https://github.com/james-6-23/codex2api (fetched 2026-08-28T04:06:01.820755+00:00, sha 7185da98e1d6)
  - homepage: https://codex2api-latest-vu8j.onrender.com (fetched 2026-08-29T10:43:09.603038+00:00, sha c9aed7800a61)
- Data as of 2026-08-30T08:39:29.467469+00:00.