Model as a Service (MaaS)

Model APIs for your applications.

Connect to selected AI models with a Cooma project key. Manage permissions, set usage limits and see what each application spends.

The Cooma Model Market showing model families, API formats and token pricing in a demo account

Model catalogue

Choose the model for the job.

Explore GLM, Kimi, DeepSeek and MiniMax alongside models from OpenAI and Anthropic. We confirm the exact version, supported APIs, token pricing and provider terms for your project.

Individual developers can request a key for an application. Business teams can set permissions and limits per project and review usage across keys.

Request current models and pricing

API integration

Make your first request.

Use your approved model ID and project key with the Responses API. Selected models also support the Messages API.

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["COOMA_API_KEY"],
    base_url="https://api.tokenfactory.cooma.ai/v1",
)

response = client.responses.create(
    model=os.environ["COOMA_MODEL_ID"],
    input="Summarise this project brief.",
)
print(response.output_text)

Access and data

Choose who can use each model.

Set model permissions, expiry dates and daily or monthly limits for each project key. Review recorded usage and charges in the portal.

For provider-hosted models, inference runs on the provider’s infrastructure. Confirm processing location, prompt retention and training-use terms for the selected model before sending sensitive information.

Discuss your requirements