As recorded at collection: Kluster.ai, 1,000 free batch requests per file

Kluster.ai is an AI cloud platform specialized in processing large volumes of requests at once using public AI models. It supports DeepSeek-R1, Llama 4 Maverick and Scout, and Qwen3. The free tier limit per file is 1,000 requests and 100MB of uploaded data. It also offers real-time processing that returns answers in under 1 second.

Sign up on the website, obtain an API key (an authentication string required to use the service), and change the API address to api.kluster.ai/v1. Since it is compatible with OpenAI, you can use your existing Python tools for OpenAI as they are, and start using the service as soon as you sign up.

Paid batch processing starts at 0.10 dollars per 1 million tokens for small models, with results returned within up to 72 hours. It is up to 50% cheaper than other services offering the same processing. Tasks that exceed the free tier limit of 1,000 requests and 100MB per file switch to paid service. Promotional credits are not always available and are offered from time to time, so you may not receive them depending on when you use the service.