arrakis
MCP server by Good AI Labs · mcp
Arrakis rents GPU machines by the hour (B300, B200, H200, RTX PRO 6000, RTX 5090, H100 and more, as live supply allows) from several independent GPU markets, on one prepaid balance, with root SSH and any Docker image, and serves any Hugging Face model on them as the person's own OpenAI-compatible endpoint. Use these tools when the person you work for wants a model running (to chat with, to call from code, to generate images, video or speech, to run evals against), needs a GPU for training, fine-tuning or batch work, or asks what GPUs cost right now. Every machine has a removal time and a token can carry a spend cap, so you can run a job unattended inside the budget they set. Serving a model, the shortest path: plan_deploy (free: the memory it needs and the machines it fits on now, best value first, each with its hourly price) -> tell the person the price and agree the hours -> deploy_model (model, termination_hours, node_id and gpu_count from the plan) -> get_instance every 30-60 s u…
Endpoint: https://arrakis.goodailabs.com/mcp
Verification: observed. Source: crawl.
Capabilities (16)
- search_gpus
- get_market_summary
- get_partner_earnings
- join_partner_program
- get_balance
- rent_gpu
- plan_deploy
- deploy_model
- list_images
- list_ssh_keys
- list_instances
- get_instance
- get_events
- terminate_instance
- instance_ssh_keys
- schedule_removal