🎼 12 Models, One API

Drop-in OpenAI replacement. Every request routed to the best model — or pick one yourself.

Default
Conductor Autoconductor

Smart routing — picks the best model for each request.

Speed
Quality
Fastest
Conductor Fastconductor-fast

Optimized for speed. Uses the fastest available model.

Speed
Quality
Best
Conductor Qualityconductor-quality

Maximum quality. Routes to the best model regardless of latency.

Speed
Quality
Creative
Conductor Creativeconductor-creative

Higher temperature, diverse outputs. Great for writing.

Speed
Quality
Google
Gemini 2.5 Flashgemini-flash

Google's fastest model. Excellent for general tasks and coding.

Speed
Quality
Google
Gemini 2.5 Progemini-pro

Google's most capable model. Deep reasoning, complex analysis.

Speed
Quality
OpenAI
GPT-4ogpt-4o

OpenAI's flagship. Strong all-rounder with vision and tool use.

Speed
Quality
OpenAI
GPT-4o Minigpt-4o-mini

Fast and cheap. Perfect for simple tasks and high-volume use.

Speed
Quality
Anthropic
Claude Sonnet 4claude-sonnet

Anthropic's balanced model. Excellent writing and reasoning.

Speed
Quality
Meta
Llama 3.3 70Bllama-70b

Meta's open-weight powerhouse. Great quality, self-hostable.

Speed
Quality
Alibaba
Qwen 2.5 7B Turboqwen-7b

Alibaba's efficient model. Fast, cheap, good for Asian languages.

Speed
Quality
DeepSeek
DeepSeek V3deepseek-v3

Top-tier reasoning at fraction of the cost. Excellent for code.

Speed
Quality

Quick Start

from openai import OpenAI

client = OpenAI(
    base_url="https://pocketclaw.tech/api/v1",
    api_key="YOUR_KEY"
)

response = client.chat.completions.create(
    model="conductor",
    messages=[{"role": "user", "content": "Hello!"}]
)
Get API KeyView Pricing