mimo2codex

https://mimodoc.chengj.online
Admin
A local LLM proxy that enables OpenAI Codex CLI/Desktop to work with any modern LLM provider like...
Tokens:30,692
Snippets:420
Trust Score:2.1
Update:4 months ago
Show doc for...
Tokens:
Raw
### Setting SenseNova API Key and Starting mimo2codex

Source: https://mimodoc.chengj.online/docs/sensenova

This example shows how to set the SENSENOVA_API_KEY environment variable and start the mimo2codex application.

```bash
export SENSENOVA_API_KEY=sk-xxxxx   # get one at https://platform.sensenova.cn/console/keys
mimo2codex
Copy
```

--------------------------------

### Initial Admin Setup

Source: https://mimodoc.chengj.online/docs/auth-deployment

Access `http://<host>:8788/admin/` in your browser after starting the service. The first user to access this page will be prompted to create the initial administrator account by providing a username, display name, and password.

```bash
open http://localhost:8788          # mac / Linux
start http://localhost:8788         # Windows
```

--------------------------------

### Docker Compose Quickstart

Source: https://mimodoc.chengj.online/docs/docker

Deploy mimo2codex using docker-compose by preparing a .env file with API keys, starting the services, and accessing the admin UI.

```bash
# 1. Prepare .env (already gitignored)
cp .env.example .env
# Edit .env, set at least one provider key:
#   MIMO_API_KEY=sk-xxxxxxxxxxxx
#   or DEEPSEEK_API_KEY=sk-xxxxxxxxxxxx

# 2. Start
docker compose up -d

# 3. Open admin UI
open http://localhost:8788          # mac/Linux
start http://localhost:8788         # Windows
Copy

```

--------------------------------

### 60-Second Start with Environment Variables

Source: https://mimodoc.chengj.online/docs/generic-providers

Use this method for a single provider setup. Set the base URL, API key, and default model via environment variables before starting mimo2codex.

```bash
export GENERIC_BASE_URL=https://dashscope.aliyuncs.com/compatible-mode/v1
export GENERIC_API_KEY=sk-your-qwen-key
export GENERIC_DEFAULT_MODEL=qwen3-max
mimo2codex --model generic

```

--------------------------------

### Starting mimo2codex with a Configured Provider

Source: https://mimodoc.chengj.online/docs/generic-providers

After setting up the providers.json file and exporting necessary API keys, start mimo2codex by specifying the desired model using its ID or shortcut.

```bash
export QWEN_API_KEY=sk-...
export KIMI_API_KEY=sk-...
mimo2codex --model qwen        # default provider = qwen

```

--------------------------------

### Install mimo2codex CLI

Source: https://mimodoc.chengj.online/

Install the mimo2codex CLI globally using npm. This provides a single binary with zero runtime configuration required.

```bash
npm install -g mimo2codex
```

--------------------------------

### Start Docker Deployment

Source: https://mimodoc.chengj.online/docs/auth-deployment

Use this command to start the Docker deployment in detached mode. Open the service in your browser using the provided URL.

```bash
docker compose up -d
open http://localhost:8788          # mac / Linux
start http://localhost:8788         # Windows
```

--------------------------------

### Recommended providers.json setup for MiniMax

Source: https://mimodoc.chengj.online/docs/minimax

Configure MiniMax as a provider in `~/.mimo2codex/providers.json`. This setup enables specific features for MiniMax compatibility. Ensure your API key is set as an environment variable.

```json
{
  "providers": [
    {
      "id": "minimax",
      "displayName": "MiniMax M2.7",
      "baseUrl": "https://api.minimaxi.com/v1",
      "envKey": "MINIMAX_API_KEY",
      "defaultModel": "MiniMax-M2.7",
      "models": [
        { "id": "MiniMax-M2.7", "contextWindow": 245760 }
      ],
      "features": {
        "minimaxCompat": true,
        "forceParallelToolCalls": true
      }
    }
  ]
}
```

--------------------------------

### Routing Fallback Example

Source: https://mimodoc.chengj.online/docs/generic-providers

Demonstrates how a client request for 'qwen3-max' falls through to the default 'mimo' provider when the configured provider ('qwen') lacks an API key.

```bash
# Client sends qwen3-max
# → byClientModel matches the qwen catalog
# → qwen has no key → fall through
# → default provider mimo → model rewritten to mimo-v2.5-pro
```

--------------------------------

### Install Pet into Codex

Source: https://mimodoc.chengj.online/docs/mimoskill

Installs a generated pet image or bundle into the Codex application. Requires the pet file and a target name.

```bash
bash mimoskill/scripts/install_pet.sh pet.png shiba
```

--------------------------------

### Generate and Install Pet (No OpenAI Key)

Source: https://mimodoc.chengj.online/docs/mimoskill

Generates a pet image and installs it into Codex without requiring an OpenAI key. For higher quality, an API key can be set.

```bash
python3 mimoskill/scripts/generate_pet.py --description "chibi shiba coder" --out pet.png
bash mimoskill/scripts/install_pet.sh pet.png shiba
# Fully quit + relaunch Codex, pick the new pet from the picker
```

--------------------------------

### Start Docker Compose Services

Source: https://mimodoc.chengj.online/docs/docker

Starts all services defined in the docker-compose.yml file in detached mode.

```bash
docker compose up -d
```

--------------------------------

### Setting up MiniMax via environment variables

Source: https://mimodoc.chengj.online/docs/minimax

A simplified setup for using MiniMax as the sole upstream provider via environment variables. Note that this method does not support the `features.minimaxCompat` switch, requiring the `providers.json` method for strict-mode sanitizers.

```bash
export GENERIC_BASE_URL=https://api.minimaxi.com/v1
export GENERIC_API_KEY=<your_key>
export GENERIC_DEFAULT_MODEL=MiniMax-M2.7
export GENERIC_FORCE_DEFAULT_MODEL=1   # key bit: unknown client models (e.g. "gpt-5.5") rewrite to MiniMax-M2.7
mimo2codex
```

--------------------------------

### Apply Codex Configuration Script (Linux/macOS)

Source: https://mimodoc.chengj.online/docs/auth-deployment

Run this script to back up existing configuration files and apply the downloaded Codex configuration to your local setup.

```bash
bash apply-*.sh
```

--------------------------------

### SenseNova Error Example

Source: https://mimodoc.chengj.online/docs/sensenova

This is an example of an error response received when wiring up SenseNova as a plain generic provider, indicating an invalid request.

```json
upstream returned 400: {"error":{"message":"Errors in message queue response",
                                  "type":"invalid_request_error","code":"3"}}
Copy
```

--------------------------------

### Windows Git Bash/WSL/Cygwin: Load Environment Variables

Source: https://mimodoc.chengj.online/docs/env-setup

For environments like Git Bash, WSL, or Cygwin on Windows, the process is the same as on macOS and Linux. Copy the example .env file and then source the shell script to load the variables.

```bash
cp .env.example .env && source scripts/load-env.sh
```

--------------------------------

### INFO model fallback applied log example

Source: https://mimodoc.chengj.online/docs/proxy-faq

Example log message indicating that an unknown client model ID was automatically mapped to a provider's default model. This is informational and the request proceeds normally.

```json
[2026-05-20T02:07:51.792Z] INFO model fallback applied — client sent unknown model id, request continues with provider default {
  provider: 'mimo',
  from: 'gpt-5.4',
  to: 'mimo-v2.5-pro',
  reason: "unknown client model — falling back to mimo provider's defaultModel"
}
```

--------------------------------

### macOS/Linux: Load Environment Variables from .env

Source: https://mimodoc.chengj.online/docs/env-setup

After copying the example .env file and editing it with your keys, use this command to load the environment variables into your current shell session. This makes the keys available to other tools launched from the same shell.

```bash
cp .env.example .env
source scripts/load-env.sh
echo $MIMO_API_KEY
mimo2codex
```

--------------------------------

### Zero-Setup OCR with MiMo Scripts

Source: https://mimodoc.chengj.online/docs/mimoskill

Performs OCR on an image using the MiMo script. It automatically selects the best available OCR engine without requiring any prior setup or API keys.

```bash
python3 mimoskill/scripts/ocr.py path/to/image.png
```

--------------------------------

### Windows PowerShell: Load Environment Variables from .env

Source: https://mimodoc.chengj.online/docs/env-setup

On Windows, use this command to copy the example .env file and then dot-source the PowerShell script to load the environment variables into your current session. Verify by echoing the variable.

```powershell
Copy-Item .env.example .env
. .\scripts\load-env.ps1
echo $env:MIMO_API_KEY
mimo2codex
```

--------------------------------

### Docker Compose Deployment

Source: https://mimodoc.chengj.online/docs/auth-deployment

The Docker image has server mode enabled by default. To start the service, use `docker compose up -d`. If you need to disable auth (not recommended), modify the `MIMO2CODEX_AUTH` environment variable in your `docker-compose.yml`.

```bash
docker compose up -d
# Just open the browser — no docker logs needed:
open http://localhost:8788          # mac / Linux
start http://localhost:8788         # Windows
```

```yaml
# docker-compose.yml
services:
  mimo2codex:
    environment:
      - MIMO2CODEX_AUTH=off
```

--------------------------------

### Stronger Docker Compose Restart

Source: https://mimodoc.chengj.online/docs/docker

Performs a full stop and then start of services, ensuring environment variables are cleanly re-injected.

```bash
docker compose down && docker compose up -d
```

--------------------------------

### Codex Runtime Override Logic

Source: https://mimodoc.chengj.online/docs/codex-enable

Explains the priority order for selecting a provider, starting with runtime overrides and falling back to defaults. Runtime overrides are checked first if a provider exists and has an API key.

```text
Pass 0: runtime override (if provider exists AND has a runtime / api key) → hit
Pass 1: user-declared generic providers (non-empty models[] in providers.json + has key)
Pass 2: built-in providers (mimo / deepseek + has key)
Fallback: defaultProviderId's defaultModel
```

--------------------------------

### Load Specific .env File with Bash

Source: https://mimodoc.chengj.online/docs/env-setup

To load environment variables from a specific file other than the default .env, use the script with the file path as an argument. This example uses Bash.

```shell
source scripts/load-env.sh .env.dev
```

--------------------------------

### Admin API: Get Codex Targets

Source: https://mimodoc.chengj.online/docs/codex-enable

Fetches a flat list of all available provider and model targets, indicating whether each has an API key configured and if it's the current override.

```http
GET /admin/api/codex-targets
```

--------------------------------

### Initialize .env File with mimo2codex

Source: https://mimodoc.chengj.online/docs/env-setup

Run this command to create the necessary `.env` and `.env.example` files in the `~/.mimo2codex/` directory. You will then need to edit the `.env` file to add your actual API keys.

```bash
mimo2codex init
```

--------------------------------

### Print Configuration for Qwen Model

Source: https://mimodoc.chengj.online/docs/generic-providers

Use this command to view the configuration snippets for the Qwen model, including auth.json and config.toml.

```bash
mimo2codex print-config --model qwen
```

--------------------------------

### Plain Docker Run - Minimal

Source: https://mimodoc.chengj.online/docs/docker

Run mimo2codex with a minimal docker command, including the API key directly in the environment variable and mapping the port. No data persistence is configured.

```bash
# Minimal (key inline, no persistence)
docker run -d --name mimo2codex \
  -p 8788:8788 \
  -e MIMO_API_KEY=sk-xxxxxxxxxxxx \
  ghcr.io/7as0nch/mimo2codex:latest
Copy

```

--------------------------------

### Get Active Runtime Override

Source: https://mimodoc.chengj.online/docs/codex-enable

Retrieves the currently active runtime override settings, if any are configured.

```APIDOC
## GET /admin/api/active-override

### Description
Retrieves the current runtime override settings. This indicates if a specific provider and model are being used at runtime, overriding the default configuration.

### Method
GET

### Endpoint
/admin/api/active-override

### Response
#### Success Response (200)
- **providerId** (string | null) - The ID of the provider currently set for runtime override, or null if no override is active.
- **modelId** (string | null) - The ID of the model currently set for runtime override, or null if no override is active.
```

--------------------------------

### Plain Docker Run - With Persistence

Source: https://mimodoc.chengj.online/docs/docker

Run mimo2codex with data persistence using a .env file for configuration, specifying a data directory inside the container, and mounting a host volume.

```bash
# With persistence + .env file
docker run -d --name mimo2codex \
  -p 8788:8788 \
  --env-file ./.env \
  -e MIMO2CODEX_DATA_DIR=/data \
  -v ~/.mimo2codex:/data \
  ghcr.io/7as0nch/mimo2codex:latest
Copy

```

--------------------------------

### Print Environment Key Configuration for Qwen Model

Source: https://mimodoc.chengj.online/docs/generic-providers

This command displays the configuration for the Qwen model when using the environment key variant, specific to the Codex CLI.

```bash
mimo2codex print-config --model qwen --env-key
```

--------------------------------

### Admin API: Get Runtime Override Status

Source: https://mimodoc.chengj.online/docs/codex-enable

Use this endpoint to retrieve the current runtime override settings. This is useful for verifying if an override is active.

```http
GET /admin/api/active-override
```

--------------------------------

### Admin API: Get Codex State

Source: https://mimodoc.chengj.online/docs/codex-enable

Retrieves a snapshot of the current Codex state, including owner, file existence, backup list, and runtime override status.

```http
GET /admin/api/codex-state
```

--------------------------------

### Offline OCR with Tesseract

Source: https://mimodoc.chengj.online/docs/mimoskill

Configures OCR to use the Tesseract engine for offline processing, suitable for environments without network access or behind firewalls. Requires Tesseract installation.

```bash
python3 mimoskill/scripts/ocr.py --engine tesseract --lang Chinese scan.png
```

--------------------------------

### Admin API: Apply Configuration

Source: https://mimodoc.chengj.online/docs/codex-enable

Applies a new configuration by writing files and creating an auto-backup. This operation replaces the entire config.toml file.

```http
POST /admin/api/codex-apply
body {providerId, modelId}
```

--------------------------------

### Load Specific .env File with PowerShell

Source: https://mimodoc.chengj.online/docs/env-setup

To load environment variables from a specific file other than the default .env, use the script with the file path as an argument. This example uses PowerShell.

```powershell
. .\scripts\load-env.ps1 .\.env.prod
```

--------------------------------

### Print CC-Switch Custom Provider Configuration for Qwen Model

Source: https://mimodoc.chengj.online/docs/generic-providers

Generates custom provider snippets for the Qwen model using the cc-switch flow.

```bash
mimo2codex print-cc-switch --model qwen
```

--------------------------------

### Docker Compose Environment Loading Order

Source: https://mimodoc.chengj.online/docs/docker

Illustrates the loading order of environment variables in Docker Compose, showing how host .env files are injected and how container-level .env files can override them.

```text
host .env
    │
    │  docker compose loads env_file → injects KEY=VALUE into container process.env
    ▼
container process starts
    │
    │  mimo2codex/cli.ts then reads <dataDir>/.env (default /data/.env)
    ▼
final process.env:
    /data/.env exists → overrides compose-injected values for matching keys
    /data/.env absent → compose-injected values stand
Copy

```

--------------------------------

### Alibaba Qwen Provider Configuration

Source: https://mimodoc.chengj.online/docs/generic-providers

Configuration for Alibaba Qwen using DashScope's OpenAI-compatible mode. Includes model details and context window sizes.

```json
{
  "id": "qwen",
  "displayName": "Qwen (DashScope)",
  "baseUrl": "https://dashscope.aliyuncs.com/compatible-mode/v1",
  "envKey": "QWEN_API_KEY",
  "defaultModel": "qwen3-max",
  "models": [
    { "id": "qwen3-max", "contextWindow": 262144 },
    { "id": "qwen3-coder-plus", "contextWindow": 1048576, "supportsReasoning": true }
  ],
  "features": { "forceParallelToolCalls": true }
}
```

--------------------------------

### Mounting Custom Data Directory in Docker

Source: https://mimodoc.chengj.online/docs/docker

Configure Docker Compose to mount a custom directory on the host to the container's data directory, allowing for data persistence and sharing across different installations.

```yaml
volumes:
  - ~/.mimo2codex:/data
Copy

```

--------------------------------

### Launch with IPv4 First DNS Resolution (Node.js)

Source: https://mimodoc.chengj.online/docs/proxy-faq

Sets the NODE_OPTIONS environment variable to prioritize IPv4 addresses during DNS resolution. Useful for mitigating DNS poisoning or IPv6 issues.

```bash
NODE_OPTIONS=--dns-result-order=ipv4first
```

--------------------------------

### MiniMax Provider Configuration with Compatibility

Source: https://mimodoc.chengj.online/docs/generic-providers

Configuration for MiniMax, including features for strict OpenAI compatibility and parallel tool calls. Use the 'minimaxCompat' preset.

```json
{
  "id": "minimax",
  "displayName": "MiniMax M2.7",
  "baseUrl": "https://api.minimaxi.com/v1",
  "envKey": "MINIMAX_API_KEY",
  "defaultModel": "MiniMax-M2.7",
  "models": [
    { "id": "MiniMax-M2.7", "contextWindow": 245760 }
  ],
  "features": {
    "minimaxCompat": true,
    "forceParallelToolCalls": true
  }
}
```

--------------------------------

### Setting Environment Variables for Authentication

Source: https://mimodoc.chengj.online/docs/auth-deployment

Demonstrates how to enable authentication for the current shell session. This is a common step when the environment variables are not being picked up correctly.

```powershell
$env:MIMO2CODEX_AUTH = "on"
```

--------------------------------

### Configuration File for Multiple Providers

Source: https://mimodoc.chengj.online/docs/generic-providers

Define multiple upstream providers in a JSON file for flexible configuration. This is the recommended approach for managing several instances.

```json
{
  "providers": [
    {
      "id": "qwen",
      "shortcut": "qwen",
      "displayName": "Qwen (DashScope)",
      "baseUrl": "https://dashscope.aliyuncs.com/compatible-mode/v1",
      "envKey": "QWEN_API_KEY",
      "defaultModel": "qwen3-max",
      "wireApi": "chat",
      "models": [
        { "id": "qwen3-max", "contextWindow": 262144 },
        { "id": "qwen3-coder-plus", "contextWindow": 1048576 }
      ],
      "features": { "forceParallelToolCalls": true }
    },
    {
      "id": "kimi",
      "shortcut": "kimi",
      "displayName": "Kimi K2",
      "baseUrl": "https://api.moonshot.cn/v1",
      "envKey": "KIMI_API_KEY",
      "defaultModel": "kimi-k2-0905-preview"
    },
    {
      "id": "ollama",
      "shortcut": "ol",
      "displayName": "Ollama (local)",
      "baseUrl": "http://127.0.0.1:11434/v1",
      "envKey": "OLLAMA_API_KEY",
      "defaultModel": "qwen2.5-coder:7b"
    },
    {
      "id": "openai-native",
      "displayName": "OpenAI (native Responses)",
      "baseUrl": "https://api.openai.com/v1",
      "envKey": "OPENAI_API_KEY",
      "defaultModel": "gpt-5",
      "wireApi": "responses"
    }
  ]
}

```

--------------------------------

### Build Local Docker Image

Source: https://mimodoc.chengj.online/docs/docker

Builds a Docker image locally from the current directory, tagged as mimo2codex:local. This is typically faster than multi-arch CI builds.

```bash
docker build -t mimo2codex:local .
```

--------------------------------

### Enable Server Mode via Shell Environment Variable

Source: https://mimodoc.chengj.online/docs/auth-deployment

Set the `MIMO2CODEX_AUTH` environment variable to `on` in your current shell session before running `npm run dev`. This setting is temporary for the session.

```bash
# PowerShell
$env:MIMO2CODEX_AUTH = "on"
npm run dev

# bash / zsh
export MIMO2CODEX_AUTH=on
npm run dev
```

--------------------------------

### Docker Compose Environment Injection Methods

Source: https://mimodoc.chengj.online/docs/docker

Compares different methods for injecting environment variables into Docker Compose services, highlighting the recommended approach for local development and CI/CD.

```text
Method| How| When| Safety  
---|---|---|---
`env_file: - .env` (compose default)| Point compose at `.env` file| Local deploy ✅| `.env` is gitignored  
`environment: - KEY=value` (hardcoded)| Literal in compose file| ❌ Don't| Key lands in git history  
`environment: - KEY=${KEY}` (from shell)| Read from invoking shell| CI / CD| OK  
`-e KEY=value` (docker run inline)| Command line| Debug| OK, but leaks to shell history  

```

--------------------------------

### Enable Server Mode via CLI Flag

Source: https://mimodoc.chengj.online/docs/auth-deployment

Use the CLI flag `--auth on` when running `npm run dev` to enable server mode for local development. This is the most direct method.

```bash
# Windows PowerShell
npm run dev -- --auth on

# macOS / Linux
npm run dev -- --auth on
```

--------------------------------

### High-Quality Image Generation with OpenAI

Source: https://mimodoc.chengj.online/docs/mimoskill

Generates an image using OpenAI's official image generation model for the best quality. Requires the PET_OPENAI_API_KEY to be set.

```bash
export PET_OPENAI_API_KEY=sk-real-openai-key
python3 mimoskill/scripts/generate_image.py --prompt "..." --out art.png
```

--------------------------------

### Free Image Generation with Pollinations

Source: https://mimodoc.chengj.online/docs/mimoskill

Generates an image using the specified prompt. It defaults to using the Pollinations provider for free generation if no OpenAI key is set.

```bash
python3 mimoskill/scripts/generate_image.py --prompt "japanese garden, watercolor, dawn" --out garden.png
```

--------------------------------

### Image Generation with Common Styles

Source: https://mimodoc.chengj.online/docs/mimoskill

Generates an image with a specified style preset, such as 'anime'. Requires the prompt and output file path.

```bash
python3 mimoskill/scripts/generate_image.py --style anime --prompt "shrine at dusk" --out shrine.png
```

--------------------------------

### Build Docker Image for Specific Platform

Source: https://mimodoc.chengj.online/docs/docker

Use this command to build a Docker image for a specific platform, such as armv7. Ensure you have the correct base image and consider compatibility notes for certain combinations like alpine + node on armv7.

```bash
docker build --platform linux/arm/v7 -t mimo2codex:armv7 .
```

--------------------------------

### Run mimo_chat.py with pollinations fallback

Source: https://mimodoc.chengj.online/docs/mimoskill

This script can be run without any API key by relying on the free pollinations fallback. It supports text and vision chat.

```bash
python3 mimoskill/scripts/mimo_chat.py "tell me a joke"

```

```bash
python3 mimoskill/scripts/mimo_chat.py --image https://x/y.png "describe this"

```

--------------------------------

### Build Docker Image Locally

Source: https://mimodoc.chengj.online/docs/docker

Build the mimo2codex Docker image locally using docker build for your current architecture. Alternatively, configure docker-compose to build from the local Dockerfile.

```bash
# Build for your current arch
docker build -t mimo2codex:local .

# Or via compose: in docker-compose.yml swap
#   image: ghcr.io/7as0nch/mimo2codex:latest   ← comment out
#   build: .                                    ← enable
docker compose build
docker compose up -d
Copy

```

--------------------------------

### Advanced MiMo Chat Configuration

Source: https://mimodoc.chengj.online/docs/mimoskill

Launches a MiMo chat session with specific model and token configurations, supporting streaming output.

```bash
python3 mimoskill/scripts/mimo_chat.py --model mimo-v2.5-pro --max-tokens 8000 --stream "long answer please"
```

--------------------------------

### Launch mimo2codex with a custom port

Source: https://mimodoc.chengj.online/docs/proxy-faq

Launch the mimo2codex application on a different port if the default is taken.

```bash
mimo2codex --port 8889
```

--------------------------------

### Generate Animated Pet Bundle

Source: https://mimodoc.chengj.online/docs/mimoskill

Generates an animated multi-state pet bundle. Specify the description and the output directory for the bundle.

```python
python3 mimoskill/scripts/generate_pet.py --description "chibi cat" --bundle ./shiba/
```

--------------------------------

### Basic MiMo Chat with Web Search

Source: https://mimodoc.chengj.online/docs/mimoskill

Initiates a chat with the MiMo model, automatically including web search capabilities. Ensure your MIMO_API_KEY is set.

```bash
export MIMO_API_KEY=sk-xxxxxxxxxxxxxxxx
python3 mimoskill/scripts/mimo_chat.py "今天上海天气"
```

--------------------------------

### Codex Configuration File (auth.json)

Source: https://mimodoc.chengj.online/docs/auth-deployment

Configure your Codex client by setting the OPENAI_API_KEY in the auth.json file. Replace the placeholder with your minted m2c key.

```json
{ "OPENAI_API_KEY": "m2c_<your token>" }
```

--------------------------------

### Enabling Authentication via CLI Flag

Source: https://mimodoc.chengj.online/docs/auth-deployment

Use the CLI flag to enable authentication, which bypasses potential issues with environment variable loading. This is a reliable method for development environments.

```bash
npm run dev -- --auth on
```

--------------------------------

### View mimo2codex Log File (Linux)

Source: https://mimodoc.chengj.online/docs/proxy-faq

Display the last 50 lines of the mimo2codex log file if it's redirected to a file. Redirects errors to /dev/null.

```bash
tail -n 50 ~/.mimo2codex/mimo2codex.log 2>/dev/null
```

--------------------------------

### View Docker Compose Logs Since a Specific Time

Source: https://mimodoc.chengj.online/docs/docker

Displays logs from the mimo2codex service generated within the last 10 minutes.

```bash
docker compose logs --since 10m mimo2codex
```

--------------------------------

### Launch mimo2codex with custom CA certificates on Windows

Source: https://mimodoc.chengj.online/docs/proxy-faq

Launch mimo2codex on Windows, specifying custom CA certificates to trust for TLS connections, often required for corporate proxies.

```powershell
$env:NODE_EXTRA_CA_CERTS = "C:\path\to\corp-ca.pem"
mimo2codex
```

--------------------------------

### Follow Docker Compose Logs

Source: https://mimodoc.chengj.online/docs/docker

Streams logs from the mimo2codex service in real-time. Useful for monitoring application output.

```bash
docker compose logs -f mimo2codex
```

--------------------------------

### Release Commands

Source: https://mimodoc.chengj.online/docs/tag-log

Commands for managing project releases via npm scripts. Use these for patch, minor, major, and beta releases.

```bash
npm run release:patch    # x.y.Z+1
npm run release:minor    # x.Y+1.0
npm run release:major    # X+1.0.0
npm run release:beta     # pre-release
```

--------------------------------

### Setting Environment Variables for Authentication (Bash)

Source: https://mimodoc.chengj.online/docs/auth-deployment

Alternative method to enable authentication for the current shell session using bash. This is useful for Linux/macOS environments.

```bash
export MIMO2CODEX_AUTH=on
```

--------------------------------

### Follow Recent Docker Compose Logs

Source: https://mimodoc.chengj.online/docs/docker

Streams the last 200 lines of logs from the mimo2codex service. Helpful for quickly reviewing recent activity.

```bash
docker compose logs -f --tail=200 mimo2codex
```

--------------------------------

### Enable Server Mode Persistently

Source: https://mimodoc.chengj.online/docs/auth-deployment

To enable server mode persistently, add `MIMO2CODEX_AUTH=on` to the `~/.mimo2codex/.env` file. Ensure the file exists by running `node dist/cli.js init` or `mimo2codex init` first.

```env
# If the file doesn't exist yet, run init once
node dist/cli.js init    # or `mimo2codex init` if you installed globally

# Edit ~/.mimo2codex/.env, add:
MIMO2CODEX_AUTH=on
```

--------------------------------

### Generate Master Key (Bash/Zsh)

Source: https://mimodoc.chengj.online/docs/auth-deployment

Use this command to generate a secure base64 encoded master key for production environments.

```bash
# bash / zsh
openssl rand -base64 32
```

--------------------------------

### Codex Enable Web UI Entry Point

Source: https://mimodoc.chengj.online/docs/codex-enable

Access the Codex Enable feature through the web UI sidebar. This is the primary entry point for managing Codex model configurations.

```text
http://127.0.0.1:8788/admin/ 
→ sidebar "Codex 启用"
```

--------------------------------

### Configure proxy for overseas users on macOS

Source: https://mimodoc.chengj.online/docs/proxy-faq

Set HTTPS and HTTP proxy environment variables for overseas users, using Clash's default port. Includes NO_PROXY configuration.

```bash
export HTTPS_PROXY=http://127.0.0.1:7890
export HTTP_PROXY=http://127.0.0.1:7890
export NO_PROXY=localhost,127.0.0.1,::1
mimo2codex
```

--------------------------------

### Setting Kimi API Key Environment Variable

Source: https://mimodoc.chengj.online/docs/kimi

This command sets the MOONSHOT_API_KEY environment variable, which is required to authenticate with the Kimi API. Obtain your API key from the Kimi platform console.

```bash
export MOONSHOT_API_KEY=sk-xxxxx   # get one at https://platform.kimi.com/console/api-keys
```

--------------------------------

### Launch mimo2codex with custom CA certificates on Mac

Source: https://mimodoc.chengj.online/docs/proxy-faq

Launch mimo2codex on macOS, specifying custom CA certificates to trust for TLS connections, often required for corporate proxies.

```bash
NODE_EXTRA_CA_CERTS=/path/to/corp-ca.pem mimo2codex
```

--------------------------------

### Configure HTTPS Proxy Temporarily (macOS)

Source: https://mimodoc.chengj.online/docs/proxy-faq

Sets environment variables for HTTPS_PROXY, HTTP_PROXY, and NO_PROXY in the current shell session. This is useful for testing proxy configurations before making them persistent.

```bash
export HTTPS_PROXY=http://127.0.0.1:7890
export HTTP_PROXY=http://127.0.0.1:7890
export NO_PROXY=localhost,127.0.0.1,::1
mimo2codex
Copy
```

--------------------------------

### Check Proxy Listener (macOS)

Source: https://mimodoc.chengj.online/docs/proxy-faq

Verifies if a process is listening on a specific TCP port using lsof. Useful for debugging proxy connection refusals.

```bash
lsof -iTCP -P | grep <port>
```

--------------------------------

### Configure proxy for China mainland users on Windows

Source: https://mimodoc.chengj.online/docs/proxy-faq

Remove proxy environment variables to connect directly when not using a proxy, suitable for China mainland users hitting MiMo or DeepSeek on Windows.

```powershell
Remove-Item env:HTTPS_PROXY -ErrorAction SilentlyContinue
Remove-Item env:HTTP_PROXY  -ErrorAction SilentlyContinue
mimo2codex
```

--------------------------------

### Codex Targets

Source: https://mimodoc.chengj.online/docs/codex-enable

Fetches a flat list of all available provider and model combinations, indicating which have API keys and which are currently set as the runtime override.

```APIDOC
## GET /admin/api/codex-targets

### Description
Provides a flat list of all pickable provider and model combinations. It indicates for each combination whether an API key is configured (`hasKey`) and if it is currently set as the active runtime override (`isCurrentOverride`).

### Method
GET

### Endpoint
/admin/api/codex-targets

### Response
#### Success Response (200)
- **targets** (array) - A list of available targets, each with:
  - **providerId** (string) - The ID of the provider.
  - **modelId** (string) - The ID of the model.
  - **hasKey** (boolean) - True if the provider has an API key configured.
  - **isCurrentOverride** (boolean) - True if this is the currently active runtime override.
```

--------------------------------

### Moonshot Kimi Provider Configuration

Source: https://mimodoc.chengj.online/docs/generic-providers

Configuration for Moonshot Kimi, including its API base URL, environment key, and model context window.

```json
{
  "id": "kimi",
  "displayName": "Kimi K2",
  "baseUrl": "https://api.moonshot.cn/v1",
  "envKey": "KIMI_API_KEY",
  "defaultModel": "kimi-k2-0905-preview",
  "models": [
    { "id": "kimi-k2-0905-preview", "contextWindow": 256000 }
  ]
}
```

--------------------------------

### Check Running Container Version via Image Labels

Source: https://mimodoc.chengj.online/docs/docker

Inspects the Docker image labels to retrieve the version information stored in 'org.opencontainers.image.version'.

```bash
docker image inspect ghcr.io/7as0nch/mimo2codex:latest \
  --format '{{index .Config.Labels "org.opencontainers.image.version"}}'
```

--------------------------------

### Configure proxy for China mainland users

Source: https://mimodoc.chengj.online/docs/proxy-faq

Unset proxy environment variables to connect directly when not using a proxy, suitable for China mainland users hitting MiMo or DeepSeek.

```bash
unset HTTPS_PROXY HTTP_PROXY
mimo2codex
```

--------------------------------

### Local Ollama / LM Studio Provider Configuration

Source: https://mimodoc.chengj.online/docs/generic-providers

Configuration for local Ollama or LM Studio instances using the open-catalog. Note that Ollama does not validate API keys.

```json
{
  "id": "ollama",
  "shortcut": "ol",
  "displayName": "Ollama (local)",
  "baseUrl": "http://127.0.0.1:11434/v1",
  "envKey": "OLLAMA_API_KEY",
  "defaultModel": "qwen2.5-coder:7b"
}
```

--------------------------------

### Apply Codex Configuration Script (Windows)

Source: https://mimodoc.chengj.online/docs/auth-deployment

Execute this PowerShell script to back up and apply the downloaded Codex configuration to your Windows machine. Ensure the execution policy is bypassed.

```powershell
powershell -ExecutionPolicy Bypass -File apply-*.ps1
```
Tokens:
Raw