API Documentation

Connect your favorite coding tools to AgentFritz using our OpenAI-compatible API.

Quick Start

1 Create an account (or log in)
2 Generate an API key from Account Settings
3 Point your tool at https://tryfritz.com/v1 with your fritz_ API key

API Endpoint

AgentFritz implements the OpenAI Chat Completions API. Any tool or SDK that supports a custom OpenAI base URL will work.

POST https://tryfritz.com/v1/chat/completions

{
  "model": "fritz-standard",
  "messages": [
    { "role": "system", "content": "You are a helpful assistant." },
    { "role": "user", "content": "Hello!" }
  ],
  "temperature": 0.7,
  "max_tokens": 4096,
  "stream": true
}

Header

Authorization: Bearer fritz_your_key_here

Aider

Easiest

Aider is an AI pair-programming tool for your terminal. It can edit code, run tests, and make git commits.

Option 1: Environment variables

export OPENAI_API_BASE=https://tryfritz.com/v1
export OPENAI_API_KEY=fritz_your_key_here
aider --model openai/fritz-standard

Option 2: Command-line flags

aider \
  --openai-api-base https://tryfritz.com/v1 \
  --openai-api-key fritz_your_key_here \
  --model openai/fritz-standard

Option 3: Config file `~/.aider.conf.yml`

openai-api-base: "https://tryfritz.com/v1"
openai-api-key: "fritz_your_key_here"
model: "openai/fritz-standard"

Continue (VS Code / JetBrains)

Continue is an open-source AI code assistant that runs inside your IDE.

YAML config `~/.continue/config.yaml`

models:
  - name: AgentFritz
    provider: openai
    model: fritz-standard
    apiBase: https://tryfritz.com/v1
    apiKey: fritz_your_key_here
    roles:
      - chat
      - edit

JSON config `~/.continue/config.json`

{
  "models": [{
    "name": "AgentFritz",
    "provider": "openai",
    "model": "fritz-standard",
    "apiBase": "https://tryfritz.com/v1",
    "apiKey": "fritz_your_key_here"
  }]
}

OpenCode

OpenCode is a terminal-based AI coding agent.

`~/.config/opencode/opencode.json`

{
  "provider": {
    "agentfritz": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "AgentFritz",
      "options": {
        "baseURL": "https://tryfritz.com/v1"
      },
      "models": {
        "fritz-standard": { "name": "Fritz Standard" }
      }
    }
  },
  "model": "agentfritz/fritz-standard"
}

`~/.local/share/opencode/auth.json`

{
  "agentfritz": {
    "type": "api",
    "key": "fritz_your_key_here"
  }
}

Claude Code

Claude Code natively uses the Anthropic protocol, so it requires a lightweight proxy wrapper to translate to the OpenAI format.

1. Install the proxy

git clone https://github.com/RichardAtCT/claude-code-openai-wrapper.git
cd claude-code-openai-wrapper
pip install -r requirements.txt

2. Configure the proxy `.env`

OPENAI_API_KEY=fritz_your_key_here
OPENAI_BASE_URL=https://tryfritz.com/v1
BIG_MODEL=fritz-standard
SMALL_MODEL=fritz-standard

3. Start the proxy, then Claude Code

# Terminal 1
python start_proxy.py

# Terminal 2
export ANTHROPIC_BASE_URL=http://localhost:8082
export ANTHROPIC_API_KEY=unused
claude

Cursor

Cursor supports custom OpenAI-compatible endpoints.

1. Open Cursor Settings → Models
2. Click Add Model
3. Set Override OpenAI Base URL to https://tryfritz.com/v1
4. Set the API key to your fritz_ key
5. Enter fritz-standard as the model name

OpenAI Python SDK

pip install openai

from openai import OpenAI

client = OpenAI(
    base_url="https://tryfritz.com/v1",
    api_key="fritz_your_key_here",
)

# Streaming
stream = client.chat.completions.create(
    model="fritz-standard",
    messages=[{"role": "user", "content": "Hello!"}],
    stream=True,
)

for chunk in stream:
    print(chunk.choices[0].delta.content or "", end="")

# Non-streaming
response = client.chat.completions.create(
    model="fritz-standard",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)

OpenAI Node.js SDK

npm install openai

import OpenAI from 'openai';

const client = new OpenAI({
  baseURL: 'https://tryfritz.com/v1',
  apiKey: 'fritz_your_key_here',
});

// Streaming
const stream = await client.chat.completions.create({
  model: 'fritz-standard',
  messages: [{ role: 'user', content: 'Hello!' }],
  stream: true,
});

for await (const chunk of stream) {
  process.stdout.write(chunk.choices[0]?.delta?.content || '');
}

// Non-streaming
const response = await client.chat.completions.create({
  model: 'fritz-standard',
  messages: [{ role: 'user', content: 'Hello!' }],
});
console.log(response.choices[0].message.content);

curl

Non-streaming

curl https://tryfritz.com/v1/chat/completions \
  -H "Authorization: Bearer fritz_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "fritz-standard",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

Streaming

curl https://tryfritz.com/v1/chat/completions \
  -H "Authorization: Bearer fritz_your_key_here" \
  -H "Content-Type: application/json" \
  -N \
  -d '{
    "model": "fritz-standard",
    "messages": [{"role": "user", "content": "Hello!"}],
    "stream": true
  }'

Available Models

Query the models endpoint to see what's available:

curl https://tryfritz.com/v1/models -H "Authorization: Bearer fritz_your_key_here"

Model	Description	Best For
fritz-standard	Fast, versatile general-purpose model	General coding, chat, analysis
fritz-reasoning	Advanced chain-of-thought reasoning model	Complex reasoning, math, architecture

Rate Limits

Rate limits depend on your plan. If you hit a limit, the API returns 429 Too Many Requests.

Plan	Price	Speed	Conversations	Messages/conv
Starter	$5/mo	Standard	Unlimited	Unlimited
Pro	$15/mo	Fast	Unlimited	Unlimited
Ultra	$30/mo	Maximum	Unlimited	Unlimited

Higher-tier plans receive faster AI response speeds and higher throughput.

Need Help?

Can't get your tool connected? Something not working?

Ask Fritz for Help