OpenAI Responses API with DeepFellow

DeepFellow supports the OpenAI Responses API at /v1/responses. If your application already uses the OpenAI SDK's responses.create() method, point it at your DeepFellow server and it will work without further changes.

Basic Usage

Replace the OpenAI base URL and API key with your DeepFellow server details.

curl -X POST 'https://deepfellow-server-host/v1/responses' \
  -H 'Authorization: Bearer DEEPFELLOW-PROJECT-API-KEY' \
  -H 'Content-Type: application/json' \
  -d '{
  "model": "qwen3:8b",
  "instructions": "You are a helpful assistant.",
  "input": "Explain how AI works."
}'

from openai import OpenAI

client = OpenAI(
    api_key="DEEPFELLOW-PROJECT-API-KEY",
    base_url="https://deepfellow-server-host/v1",
)

response = client.responses.create(
    model="qwen3:8b",
    instructions="You are a helpful assistant.",
    input="Explain how AI works.",
)

print(response.output_text)

import OpenAI from 'openai';

const client = new OpenAI({
    apiKey: 'DEEPFELLOW-PROJECT-API-KEY',
    baseURL: 'https://deepfellow-server-host/v1'
});

const response = await client.responses.create({
    model: 'qwen3:8b',
    instructions: 'You are a helpful assistant.',
    input: 'Explain how AI works.'
});

console.log(response.output_text);

Tools

The /v1/responses endpoint supports function calls, MCP servers, file search, image generation, and more. See Using Tools for full examples and configuration details.

Current Limitations

The following parameters are accepted for API compatibility but are not yet active:

• stream — only false is supported; streaming is not yet available
• previous_response_id — multi-turn state via response chaining is not yet supported
• tool_choice — the model always uses auto selection regardless of this value
• service_tier — has no effect; included for compatibility

Related Pages

• Secure OpenAI Inference with DeepFellow — use /v1/chat/completions with OpenAI models through DeepFellow's anonymization layer
• API Reference: POST /v1/responses — full parameter and response schema reference