codex-lb 1.6.3

Codex load balancer and proxy for ChatGPT accounts with usage dashboard

codex-lb

Load balancer for ChatGPT accounts. Pool multiple accounts, track usage, manage API keys, view everything in a dashboard.

More screenshots

Settings	Login

Dashboard (dark)	Accounts (dark)	Settings (dark)

Features

Account Pooling Load balance across multiple ChatGPT accounts	Usage Tracking Per-account tokens, cost, 28-day trends	API Keys Per-key rate limits by token, cost, window, model
Dashboard Auth Password + optional TOTP	OpenAI-compatible Codex CLI, OpenCode, any OpenAI client	Auto Model Sync Available models fetched from upstream

Quick Start

# Docker (recommended)
docker volume create codex-lb-data
docker run -d --name codex-lb \
  -p 2455:2455 -p 1455:1455 \
  -v codex-lb-data:/var/lib/codex-lb \
  ghcr.io/soju06/codex-lb:latest

# or uvx
uvx codex-lb

Open localhost:2455 → Add account → Done.

Client Setup

Point any OpenAI-compatible client at codex-lb. If API key auth is enabled, pass a key from the dashboard as a Bearer token.

Logo	Client	Endpoint	Config
	Codex CLI	http://127.0.0.1:2455/backend-api/codex	~/.codex/config.toml
	OpenCode	http://127.0.0.1:2455/v1	~/.config/opencode/opencode.json
	OpenClaw	http://127.0.0.1:2455/v1	~/.openclaw/openclaw.json
	OpenAI Python SDK	http://127.0.0.1:2455/v1	Code

 Codex CLI / IDE Extension

~/.codex/config.toml:

model = "gpt-5.3-codex"
model_reasoning_effort = "xhigh"
model_provider = "codex-lb"

[model_providers.codex-lb]
name = "OpenAI"  # required — enables remote /responses/compact
base_url = "http://127.0.0.1:2455/backend-api/codex"
wire_api = "responses"
supports_websockets = true

Optional: enable native upstream WebSockets for Codex streaming while keeping codex-lb pooling:

export CODEX_LB_UPSTREAM_STREAM_TRANSPORT=websocket

auto is the default and uses native WebSockets for native Codex headers or models that prefer them. You can also switch this in the dashboard under Settings -> Routing -> Upstream stream transport.

Note: Codex itself does not currently expose a stable documented wire_api = "websocket" provider mode. If you want to experiment on the Codex side, the current CLI exposes under-development feature flags:

[features]
responses_websockets = true
# or
responses_websockets_v2 = true

These flags are experimental and do not replace wire_api = "responses".

If upstream websocket handshakes must use environment proxies in your deployment, set CODEX_LB_UPSTREAM_WEBSOCKET_TRUST_ENV=true. By default websocket handshakes connect directly to match Codex CLI's native transport.

With API key auth:

[model_providers.codex-lb]
name = "OpenAI"
base_url = "http://127.0.0.1:2455/backend-api/codex"
wire_api = "responses"
env_key = "CODEX_LB_API_KEY"
supports_websockets = true

export CODEX_LB_API_KEY="sk-clb-..."   # key from dashboard
codex

Verify WebSocket transport

Use a one-off debug run:

RUST_LOG=debug codex exec "Reply with OK only."

Healthy websocket signals:

• CLI logs contain connecting to websocket and successfully connected to websocket
• codex-lb logs show WebSocket /backend-api/codex/responses
• codex-lb logs do not show fallback POST /backend-api/codex/responses for the same run

If you run codex-lb behind a reverse proxy, make sure it forwards WebSocket upgrades.

Migrating from direct OpenAI — codex resume filters by model_provider; old sessions won't appear until you re-tag them:

# JSONL session files (all versions)
find ~/.codex/sessions -name '*.jsonl' \
  -exec sed -i '' 's/"model_provider":"openai"/"model_provider":"codex-lb"/g' {} +

# SQLite state DB (>= v0.105.0, creates ~/.codex/state_*.sqlite)
sqlite3 ~/.codex/state_5.sqlite \
  "UPDATE threads SET model_provider = 'codex-lb' WHERE model_provider = 'openai';"

 OpenCode

Important: Use the built-in openai provider with baseURL override — not a custom provider with @ai-sdk/openai-compatible. Custom providers use the Chat Completions API which drops reasoning/thinking content. The built-in openai provider uses the Responses API, which properly preserves encrypted_content and multi-turn reasoning state.

~/.config/opencode/opencode.json:

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "openai": {
      "options": {
        "baseURL": "http://127.0.0.1:2455/v1",
        "apiKey": "{env:CODEX_LB_API_KEY}"
      },
      "models": {
        "gpt-5.4": {
          "name": "GPT-5.4",
          "reasoning": true,
          "options": { "reasoningEffort": "high", "reasoningSummary": "detailed" },
          "limit": { "context": 1050000, "output": 128000 }
        },
        "gpt-5.3-codex": {
          "name": "GPT-5.3 Codex",
          "reasoning": true,
          "options": { "reasoningEffort": "high", "reasoningSummary": "detailed" },
          "limit": { "context": 272000, "output": 65536 }
        },
        "gpt-5.1-codex-mini": {
          "name": "GPT-5.1 Codex Mini",
          "reasoning": true,
          "options": { "reasoningEffort": "high", "reasoningSummary": "detailed" },
          "limit": { "context": 272000, "output": 65536 }
        },
        "gpt-5.3-codex-spark": {
          "name": "GPT-5.3 Codex Spark",
          "reasoning": true,
          "options": { "reasoningEffort": "xhigh", "reasoningSummary": "detailed" },
          "limit": { "context": 128000, "output": 65536 }
        }
      }
    }
  },
  "model": "openai/gpt-5.3-codex"
}

This overrides the built-in openai provider's endpoint to point at codex-lb while keeping the Responses API code path that handles reasoning properly.

export CODEX_LB_API_KEY="sk-clb-..."   # key from dashboard
opencode

 OpenClaw

~/.openclaw/openclaw.json:

{
  "agents": {
    "defaults": {
      "model": { "primary": "codex-lb/gpt-5.3-codex" }
    }
  },
  "models": {
    "mode": "merge",
    "providers": {
      "codex-lb": {
        "baseUrl": "http://127.0.0.1:2455/v1",
        "apiKey": "${CODEX_LB_API_KEY}",   // or "dummy" if API key auth is disabled
        "api": "openai-completions",
        "models": [
          { "id": "gpt-5.3-codex", "name": "GPT-5.3 Codex" },
          { "id": "gpt-5.3-codex-spark", "name": "GPT-5.3 Codex Spark" }
        ]
      }
    }
  }
}

Set the env var or replace ${CODEX_LB_API_KEY} with a key from the dashboard. If API key auth is disabled, any value works.

 OpenAI Python SDK

from openai import OpenAI

client = OpenAI(
    base_url="http://127.0.0.1:2455/v1",
    api_key="sk-clb-...",  # from dashboard, or any string if auth is disabled
)

response = client.chat.completions.create(
    model="gpt-5.3-codex",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)

API Key Authentication

API key auth is disabled by default — the proxy is open to any client. Enable it in Settings → API Key Auth on the dashboard.

When enabled, clients must pass a valid API key as a Bearer token:

Authorization: Bearer sk-clb-...

Creating keys: Dashboard → API Keys → Create. The full key is shown only once at creation. Keys support optional expiration, model restrictions, and rate limits (tokens / cost per day / week / month).

Configuration

Environment variables with CODEX_LB_ prefix or .env.local. See .env.example. Dashboard auth is configured in Settings. SQLite is the default database backend; PostgreSQL is optional via CODEX_LB_DATABASE_URL (for example postgresql+asyncpg://...).

Data

Environment	Path
Local / uvx	~/.codex-lb/
Docker	/var/lib/codex-lb/

Backup this directory to preserve your data.

Development

# Docker
docker compose watch

# Local
uv sync && cd frontend && bun install && cd ..
uv run fastapi run app/main.py --reload        # backend :2455
cd frontend && bun run dev                     # frontend :5173

Contributors ✨

Thanks goes to these wonderful people (emoji key):

Soju06 💻 ⚠️ 🚧 🚇	Jonas Kamsker 💻 🐛 🚧	Quack 💻 🐛 🚧 🎨	Jill Kok, San Mou 💻 ⚠️ 🚧 🐛	PARK CHANYOUNG 📖 💻 ⚠️	Choi138 💻 🐛 ⚠️	LYA⚚CAP⚚OCEAN 💻 ⚠️
Eugene Korekin 💻 🐛 ⚠️	jordan 💻 🐛 ⚠️	DOCaCola 🐛 ⚠️ 📖	JoeBlack2k 💻 🐛 ⚠️	Peter A. 📖	Hannah Markfort 💻 ⚠️	mws-weekend-projects 💻 ⚠️
Quang Do 💻 ⚠️	Anand Aiyer 🐛 💻 ⚠️	defin85 💻 🐛 ⚠️	Jacky Fong 💻 🐛	flokosti96 💻 ⚠️	Woonggi Min 💻 ⚠️

This project follows the all-contributors specification. Contributions of any kind welcome!