instructor-ainative 0.1.0

Pre-configured instructor client for structured output extraction via AINative's free LLM API

instructor-ainative

Pre-configured instructor client for structured output extraction using AINative's free LLM API.

Zero setup. Works with Llama 3.3 70B, Qwen3, DeepSeek 4, and Kimi K2 — all free.

Install

pip install instructor-ainative

Quick Start

from instructor_ainative import get_client
from pydantic import BaseModel

class User(BaseModel):
    name: str
    age: int

client = get_client()  # Auto-provisions free API key
user = client.chat.completions.create(
    model="meta-llama/Llama-3.3-70B-Instruct",
    response_model=User,
    messages=[{"role": "user", "content": "Extract: John is 30"}],
)
print(user)  # User(name='John', age=30)

How It Works

instructor-ainative wraps instructor with a pre-configured OpenAI client pointing at AINative's free, OpenAI-compatible API. You get structured Pydantic output extraction from open-source models without any API key setup.

On first use, the package auto-provisions a free API key (72-hour TTL). Claim your account at ainative.studio/signup for permanent access.

Available Models

Use aliases or full model IDs:

from instructor_ainative import get_model, MODELS

# Aliases
get_model("llama")     # meta-llama/Llama-3.3-70B-Instruct
get_model("qwen")      # qwen3-coder-flash
get_model("deepseek")  # deepseek-4-flash
get_model("kimi")      # kimi-k2

Async Support

import asyncio
from instructor_ainative import get_async_client
from pydantic import BaseModel

class User(BaseModel):
    name: str
    age: int

async def main():
    client = get_async_client()
    user = await client.chat.completions.create(
        model="meta-llama/Llama-3.3-70B-Instruct",
        response_model=User,
        messages=[{"role": "user", "content": "Extract: John is 30"}],
    )
    print(user)

asyncio.run(main())

Configuration

API Key Resolution Order

1. Explicit api_key parameter
2. AINATIVE_API_KEY environment variable
3. ZERODB_API_KEY environment variable
4. ~/.zerodb/credentials.json (shared with ZeroDB ecosystem)
5. Auto-provision (free, 72-hour TTL)

Bring Your Own Key

export AINATIVE_API_KEY=your-key-here

# Or pass directly
client = get_client(api_key="your-key-here")

Custom Base URL

client = get_client(base_url="https://your-proxy.example.com/v1")

Instructor Mode

import instructor
client = get_client(mode=instructor.Mode.TOOLS)  # Default is JSON mode

Complex Extraction

from instructor_ainative import get_client
from pydantic import BaseModel, Field
from typing import List

class Address(BaseModel):
    street: str
    city: str
    state: str
    zip_code: str = Field(description="5-digit ZIP code")

class Contact(BaseModel):
    name: str
    email: str
    phone: str
    addresses: List[Address]

client = get_client()
contact = client.chat.completions.create(
    model="meta-llama/Llama-3.3-70B-Instruct",
    response_model=Contact,
    messages=[{
        "role": "user",
        "content": """
        Extract contact info: Jane Smith, [e-mail], 555-0123.
        Lives at 123 Main St, Austin TX 78701 and
        456 Oak Ave, Denver CO 80202.
        """
    }],
)

Why instructor-ainative?

Feature	instructor + OpenAI	instructor-ainative
Setup	Get API key, set env var	pip install and go
Cost	Pay per token	Free
Models	GPT-4o, etc.	Llama 70B, Qwen, DeepSeek
Structured output	Yes	Yes
Auto-provisioning	No	Yes

License

MIT