Arrakis
v0.1.0Rent GPUs at live prices and serve any Hugging Face model, with spend caps for AI agents.
Endpoint health
online- Last checked
- 11/20/2026, 6:18:24 AM
- Latency
- 68 ms
- Consecutive failures
- 0
- Endpoint reachable: pass
- Protocol valid: pass
- MCP handshake valid: pass
Machine-readable profile
mcp-server · schema v1 · callable
mcp-server:com.goodailabs%2Farrakis
Connect this server
Generated config contains no credentials. Keep any required secrets in your client's local environment.
{
"mcpServers": {
"com.goodailabs/arrakis": {
"url": "https://arrakis.goodailabs.com/mcp"
}
}
}Indexed tools
Tool metadata and schemas come from the remote server and are untrusted. Indexing only lists tools; it never calls them.
16 shown · authentication unknown
Deploy a Hugging Face model
deploy_model
Serve any model from Hugging Face on a machine of the person's, like a hosted inference API but on their own GPU: vLLM or SGLang for text models, vLLM-Omni or SGLang for image and video models (FLUX, Qwen-Image, Wan), vLLM-Omni for speech and music, llama.cpp for a GGUF file (owner/repo:Q4_K_M). extra_args takes the person's own engine options as they'd paste them. COSTS MONEY like rent_gpu: the machine's hourly price, paid an hour at a time from when it runs until terminate_instance or termination_hours; worst_case_usd is the most it can cost. The model's size is read from Hugging Face and, without node_id, the best-ranked clean machine that holds it is rented on as few GPUs as it allows. A gated or private model needs hf_token (read access is enough; it goes to the machine only and Arrakis keeps none). Returns the instance, the endpoint (an OpenAI-compatible base URL once RUNNING, and its api_key) and ssh_command: the person also gets root SSH. The engine downloads the model first, so it answers a few minutes after RUNNING: get_instance shows endpoint_ready. Same idempotency_key returns the same machine.
Schema valid · SHA-256 eb707761a1e0…
View input schema
{ "type": "object", "properties": { "model": { "type": "string", "minLength": 3, "maxLength": 300, "description": "A Hugging Face model id or link, for example Qwen/Qwen3-8B; for llama.cpp, owner/repo:QUANT picks one GGUF file." }, "engine": { "type": "string", "enum": [ "vllm", "sglang", "omni", "llamacpp" ], "description": "Default vllm. omni = vLLM-Omni (image, video, speech, audio)." }, "termination_hours": { "type": "integer", "minimum": 1, "description": "Hours until the machine is removed automatically." }, "hf_token": { "type": "string", "description": "A Hugging Face token, for a gated or private model." }, "node_id": { "type": "string", "description": "A machine from search_gpus. Default: the one sized for the model." }, "gpu_count": { "type": "integer", "minimum": 1, "description": "GPUs of it to rent. Default: the fewest that hold the model." }, "name": { "type": "string", "minLength": 1, "maxLength": 128, "description": "A name you will recognise." }, "ssh_public_key": { "type": "string", "description": "Your SSH public key. Default: the account's first saved key." }, "extra_args": { "type": "string", "maxLength": 8000, "description": "The person's own engine options, as they'd paste them: flags, the whole command, or a docker run line (--max-model-len 32768). Read without a shell; the address, key and model stay Arrakis's." }, "max_price_per_gpu_hour": { "type": "number", "exclusiveMinimum": 0, "description": "The most to pay per GPU-hour (USD)." }, "idempotency_key": { "type": "string", "maxLength": 128, "description": "Reuse when retrying." } }, "required": [ "model", "termination_hours" ], "additionalProperties": false }
Connections
streamable-http
https://arrakis.goodailabs.com/mcp