anoman-ai

Official Python SDK for the Anoman AI gateway. Async-first, OpenAI-compatible, with batch polling and streaming support.

Installation

pip install anoman-ai

Quick Start

from anoman_ai import AnomanAI

async with AnomanAI(api_key="anm-sk-your-key-here") as client:
    response = await client.chat_completion(
        model="claude-sonnet",
        messages=[{"role": "user", "content": "Explain quantum computing in one sentence."}],
    )
    print(response.choices[0].message.content)

Streaming

async for chunk in client.chat_completion_stream(
    model="claude-sonnet",
    messages=[{"role": "user", "content": "Write a haiku about AI safety."}],
):
    print(chunk.choices[0].delta.content, end="")

Batch Routing

Anoman can batch-route non-interactive requests for lower cost:

# If the gateway returns a 202, you get a batch job
# Use poll_batch() to wait for the result
result = await client.poll_batch("job_abc123", max_wait_seconds=300)
print(result.result.choices[0].message.content)
print(f"Saved ${result.savings_usd}")

Embeddings

response = await client.create_embedding(
    model="text-embedding-3-small",
    input="Hello world",
)
print(response.data[0].embedding)

Models

models = await client.list_models()
for model in models.data:
    print(f"{model.id} ({model.owned_by})")

Error Handling

import httpx

try:
    response = await client.chat_completion(
        model="claude-sonnet",
        messages=[{"role": "user", "content": "Hello"}],
    )
except httpx.HTTPStatusError as e:
    if e.response.status_code == 401:
        print("Invalid API key")
    elif e.response.status_code == 403:
        print("Blocked by guardrail")
    elif e.response.status_code == 429:
        print("Rate limited")

Configuration

client = AnomanAI(
    api_key="anm-sk-...",
    base_url="https://api.anoman.io",  # default
    timeout=60.0,                       # 60s default
)

Documentation

Full API documentation: https://docs.anoman.io