Features
- Shared GPU inference cluster for open-source AI models, running 24/7
- Personal OpenAI-compatible API key
- Models for language, reasoning, vision, audio, embedding, reranking, text-to-speech and speech-to-text
- Includes DeepSeek V4-Flash, Gemma 4, Qwen 3.6, Kokoro and Whisper
- No token limits on the core pool models
- Private Discord community with events, workshops and hackathons
- Quarterly member votes on models to add
- Data processed in the EU with no logging of prompts or responses
Pricing Details
The tool uses a pricing model based on monthly subscriptions, paid memberships, and premium subscriptions; there is no free plan listed. Membership starts with the nan_community plan at 14.99 euros per month, including tax, and provides access to Discord, member-only channels, live discussions, events, workshops, and hackathons, as well as a spot on the waiting list for access to inference. The nan_member plan costs 70 euros per month, including tax, and provides access to the shared GPU pool and open models, DeepSeek V4-Flash with a limit of 2 billion tokens per month, and a personal OpenAI-compatible API key, along with community benefits and the ability to vote on models. Meanwhile, the GLM 5.2 Premium plan costs 200 euros per month, including tax, and includes all the benefits of the nan_member plan, plus access to GLM 5.2 with a limit of 3 billion tokens per billing period, 400 million tokens per 4-hour rolling window, a context size of up to 500,000 tokens, and 5 concurrent requests. Both the nan_member and GLM 5.2 Premium plans are available subject to seat availability, with monthly subscriptions that can be canceled at any time without a long-term commitment.