Freebuff2API

Freebuff2API is an OpenAI-compatible proxy server for Freebuff. It translates standard OpenAI API requests into Freebuff's backend format, enabling the use of Freebuff's free models with any OpenAI-compatible client, SDK, or CLI tool.

Key features include:

• OpenAI Compatible API: Works with standard OpenAI endpoints.
• Stealth Request Handling: Uses dynamic, randomized client fingerprints to mimic official Freebuff SDK behavior.
• Multi-Token Rotation: Automatically cycles through multiple auth tokens at a configurable interval.
• HTTP Proxy Support: Allows routing outbound traffic through an upstream proxy.

Freebuff2API is an OpenAI-compatible proxy server for Freebuff. It converts standard OpenAI API requests into the Freebuff backend format, allowing you to use Freebuff's free models with any OpenAI-compatible client, SDK, or command-line tool.

Core Features:

• OpenAI Compatible API: Standard endpoints that work out-of-the-box with compatible clients.
• High Stealth Request Handling: Uses dynamic random client identifiers to simulate official Freebuff SDK behavior.
• Multi-Token Rotation: Supports multiple authentication tokens with a built-in automatic rotation mechanism.
• HTTP Proxy Support: Allows configuring an upstream HTTP proxy for all external requests.

To build the binary manually, ensure you have Go 1.23+ installed.

git clone https://github.com/Quorinex/Freebuff2API.git
cd Freebuff2API
go build -o Freebuff2API .
./Freebuff2API -config config.json

To build from source, ensure you have Go 1.23+ installed.

git clone https://github.com/Quorinex/Freebuff2API.git
cd Freebuff2API
go build -o Freebuff2API .
./Freebuff2API -config config.json

Freebuff2API requires at least one Freebuff Auth Token. You can obtain tokens using one of two methods:

Method 1: Web (Recommended)

1. Visit https://freebuff.llm.pm.
2. Log in with your Freebuff account.
3. Your Auth Token will be displayed on the page. Copy this value to use as AUTH_TOKENS.

Method 2: Freebuff CLI

1. Install the CLI globally:

   npm i -g freebuff

2. Run freebuff in your terminal and follow the login prompts.
3. Locate the authToken in your local credentials file:
   • Windows: C:\\Users\\<Username>>>\.config\\manicode\\credentials.json
   • Linux / macOS: ~/.config/manicode/credentials.json

Tip: You can log in with multiple accounts and provide all tokens to increase concurrent throughput.

You can run Freebuff2API using pre-built multi-arch images from GHCR or by building the image manually.

Run pre-built image:

docker run -d --name Freebuff2API \
  -p 8080:8080 \
  -e AUTH_TOKENS="token1,token2" \
  ghcr.io/quorinex/freebuff2api:latest

Manual build and run:

docker build -t Freebuff2API .
docker run -d -p 8080:8080 -e AUTH_TOKENS="token1,token2" Freebuff2API

Freebuff2API requires one or more Freebuff auth tokens. You can obtain them using one of two methods:

Method 1: Web (Recommended)

1. Visit https://freebuff.llm.pm.
2. Log in with your Freebuff account.
3. Copy the auth token displayed on the page.

Method 2: Freebuff CLI

1. Install the CLI: npm i -g freebuff.
2. Run freebuff and follow the login prompts.
3. Locate your token in the local credentials file:
   • Windows: C:\\Users\\<username>\\.config\\manicode\\credentials.json
   • Linux / macOS: ~/.config/manicode/credentials.json
4. Extract the authToken value from the JSON file.

{
  "default": {
    "id": "user_10293847",
    "name": "Zhang San",
    "email": "zhangsan@example.com",
    "authToken": "fa82b5c1-e39d-4c7a-961f-d2b3c4e5f6a7",
    ...
  }
}

You can run Freebuff2API using pre-built multi-arch images from GHCR or by building from source.

Run pre-built image

docker run -d --name Freebuff2API \
  -p 8080:8080 \
  -e AUTH_TOKENS="token1,token2" \
  ghcr.io/quorinex/freebuff2api:latest

Build and run from source

docker build -t Freebuff2API .
docker run -d -p 8080:8080 -e AUTH_TOKENS="token1,token2" Freebuff2API

To ensure compatibility with upstream providers that may have stricter JSON Schema requirements, the server performs Tool Schema Normalization.

When a request contains a tools array, the server:

1. Resolves local $ref pointers (supporting both #/definitions/ and #/$defs/).
2. Simplifies anyOf and oneOf combinators by removing null types.
3. Flattens type arrays into a single primitive type (preferring the first non-null type).
4. Cleans up enum, const, and nullable fields to match a conservative subset of JSON Schema.

This allows clients like LobeChat, which emit complex schemas, to work seamlessly with the proxy.

If the server is configured with API keys, you must provide one of the following in your request headers:

1. Custom Header: x-api-key: <YOUR_API_KEY>
2. Bearer Token: Authorization: Bearer <YOUR_API_KEY>