Ross ROSS = Recommend OSS · open-source software intelligence for agents

lidge-jun/opencodex

Universal provider proxy for OpenAI Codex & Claude Code — use any LLM (Claude, Gemini, Grok, DeepSeek, Ollama…) with Codex CLI, App, SDK, and Claude Code observed · 2026-08-28

github.com/lidge-jun/opencodex · homepage · TypeScript · MIT (permissive) observed · 2026-08-28

Health v2 · maintenance only

76/100

  • Activity 99
  • Release rhythm 87
  • Longevity 5

Flags: young

How is this computed?

round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10) — computed 2026-09-03. Adoption (stars, forks) is never an input.

  • gap_med: 0
  • age_days: 76
  • days_rel: 8
  • days_push: 7
  • n_releases_24m: 152

Full methodology

Adoption not part of the score

12224 stars · 898 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

opencodex is a lightweight local proxy (installed via npm, running on a bundled Bun runtime) that translates OpenAI Codex's Responses API and Claude Code's wire format into any LLM provider's API — Anthropic, Gemini, Grok, DeepSeek, Kimi, Ollama, OpenRouter, and 40+ others. It supports streaming, tool calls, reasoning tokens, and images in both directions, with a web dashboard, model routing, ChatGPT account pooling, and sub-agent support, all without patching the client.

Use cases

  • use Claude or Gemini models with OpenAI Codex CLI
  • run Claude Code on non-Anthropic models
  • route Codex App requests to DeepSeek or Ollama
  • pool multiple ChatGPT accounts for Codex quota balancing
  • translate OpenAI Responses API to Anthropic Messages format
  • add web search and vision sidecars to routed models
  • pin multiple models as Codex spawn_agent sub-agents

When to choose

  • you use Codex CLI/App/SDK or Claude Code but want models from other providers
  • you want a zero-patch local proxy that handles streaming, tool calls, and reasoning tokens
  • you need ChatGPT account pooling with quota-aware routing
  • you want routed models to appear natively in the Codex model picker

When to avoid

  • you only use OpenAI models directly with no need for provider translation
  • you need a hosted/production gateway rather than a local developer proxy
  • your client is not Codex, Claude Code, Claude Desktop, or Grok Build
  • you cannot run Node >= 18 or install global npm packages

Facets

cli-tool · maturity active

proxy llm-inference api-framework http-server cli middleware large-language-models developer-tools artificial-intelligence apis cli cross-platform windows llm-proxy codex-cli claude-code api-translation model-routing provider-adapter openai-responses-api anthropic gemini ollama openrouter chatgpt-account-pool sub-agents web-dashboard ai-agents command-line nodejs macos linux

6 sources

Member repositories

RepositoryRoleHealth v2
lidge-jun/opencodexmain76

For agents

markdown · JSON · MCP: product_card(name="lidge-jun/opencodex")

Data as of 2026-08-30T08:39:29.467469+00:00 · Report a problem