gpu.nz GPU cloud
Rent cloud GPUs from AI agents: live prices, VRAM estimates, launch, logs and stop.
사용해야 할까요
품질 및 안전성
발견 사항 (5)
- HIGHTool poisoning patterns detected
- LOWTool 'get_instance' description lacks action verbget_instance에서
- LOWTool description contains role marker that could confuse chat modelsget_account에서
- LOWTool description contains role marker that could confuse chat modelslist_ssh_keys에서
- LOWTool description contains role marker that could confuse chat modelsadd_ssh_key에서
도구 정의와 프로토콜 준수에 대한 자동 분석을 기반으로 합니다.
컨텍스트 비용
이는 서버의 도구가 모델의 컨텍스트에 로드될 때마다 소비되는 대략적인 토큰 수입니다. 수치가 높을수록 다른 작업에 사용할 수 있는 주의가 줄어듭니다.
설치
원클릭 설치
`claude_desktop_config.json` 파일에 다음을 추가하세요:
{
"mcpServers": {
"gpu-cloud": {
"url": "https://gpu.nz/mcp"
}
}
}원격 엔드포인트
https://gpu.nz/mcpstreamable-http할 수 있는 일
도구 목록
도구 (16)
🟢list_gpu_models(min_vram_gb, query)
GPU models available on gpu.nz right now with VRAM, class and the cheapest hourly price per GPU. Use it to pick a gpu_slug for search_offers.
입력 스키마
{
"type": "object",
"properties": {
"min_vram_gb": {
"description": "Only models with at least this much VRAM per GPU",
"type": "number",
"minimum": 0
},
"query": {
"description": "Filter by name, e.g. \"4090\" or \"H100\"",
"type": "string",
"maxLength": 40
}
},
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢search_offers(gpu, num_gpus, min_vram_gb, max_price_rub_h, country, ...)
Search rentable GPU servers (offers) with live prices. Returns the cheapest matches first unless `order` says otherwise. Each result has an offer id for create_instance.
입력 스키마
{
"type": "object",
"properties": {
"gpu": {
"description": "GPU model slug(s), comma-separated, e.g. \"rtx-4090-24gb\" or \"h100-sxm-80gb,h100-pcie-80gb\" (see list_gpu_models)",
"type": "string",
"maxLength": 200
},
"num_gpus": {
"description": "GPUs per server, comma-separated, e.g. \"1\" or \"2,4,8\"",
"type": "string",
"maxLength": 20
},
"min_vram_gb": {
"description": "Minimum VRAM per GPU, GB",
"type": "number",
"minimum": 0
},
"max_price_rub_h": {
"description": "Maximum price for the whole server, rubles per hour",
"type": "number",
"minimum": 0
},
"country": {
"description": "ISO-2 country codes, comma-separated, e.g. \"DE,NL,FI\"",
"type": "string",
"maxLength": 60
},
"min_reliability": {
"description": "Minimum host reliability, 0–1 (0.98 is a good bar)",
"type": "number",
"minimum": 0,
"maximum": 1
},
"verified": {
"description": "Only verified datacenters",
"type": "boolean"
},
"min_disk_gb": {
"description": "Minimum disk the server can give, GB",
"type": "number",
"minimum": 0
},
"type": {
"description": "on_demand (default) or interruptible (cheaper, can be preempted)",
"type": "string",
"enum": [
"on_demand",
"interruptible"
]
},
"order": {
"description": "Sort: price (default), dlperf_per_price for value, reliability, vram, inet_down",
"type": "string",
"enum": [
"price",
"-price",
"dlperf",
"dlperf_per_price",
"reliability",
"vram",
"inet_down"
]
},
"limit": {
"description": "How many offers to return (default 10)",
"type": "integer",
"minimum": 1,
"maximum": 30
}
},
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢get_offer(offer_id)
Details and the current price of one offer by id (from search_offers).
입력 스키마
{
"type": "object",
"properties": {
"offer_id": {
"type": "string",
"minLength": 1,
"maxLength": 64,
"description": "Offer id"
}
},
"required": [
"offer_id"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢list_templates
Ready environments a server can start with: ComfyUI, Stable Diffusion Forge, Ollama + Open WebUI, vLLM, Jupyter/PyTorch, Kohya, FluxGym, Unsloth, chats with preloaded models, plain Ubuntu + CUDA, or a custom Docker image. Pass `slug` as `template` to create_instance.
입력 스키마
{
"type": "object",
"properties": {},
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢estimate_vram(model, precision, context_tokens, concurrent_requests)
How much GPU memory an AI model needs (LLMs such as Qwen, DeepSeek, Llama, Gemma, GPT-OSS; image and video models such as FLUX, SDXL, Wan) at a given precision and context length, and which GPUs on gpu.nz fit it with their current price. Call without `model` to list the known models.
입력 스키마
{
"type": "object",
"properties": {
"model": {
"description": "Model name or id, e.g. \"qwen3-32b\", \"Llama 3.3 70B\", \"flux.1-dev\"",
"type": "string",
"maxLength": 80
},
"precision": {
"description": "Weights precision: \"fp16\", \"fp8\", \"q8\", \"q6\", \"q5\", \"q4\" (LLMs; default q4 for LLMs, fp16 for image models)",
"type": "string",
"maxLength": 20
},
"context_tokens": {
"description": "LLM context length in tokens (default 8192)",
"type": "integer",
"minimum": 512,
"maximum": 1048576
},
"concurrent_requests": {
"description": "Parallel sequences sharing the GPU (default 1)",
"type": "integer",
"minimum": 1,
"maximum": 256
}
},
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢get_account
The signed-in account and its balance, hourly burn and runway. Needs a key with user:read and billing:read.
입력 스키마
{
"type": "object",
"properties": {},
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢list_instances(status)
Your rented servers with status, GPU, price and SSH command. Needs instances:read.
입력 스키마
{
"type": "object",
"properties": {
"status": {
"description": "active (default): servers that are not destroyed; all: include destroyed ones",
"type": "string",
"enum": [
"active",
"all"
]
}
},
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢get_instance(instance_id, include_password)
Status, SSH command, app URLs (ComfyUI, Jupyter, chat…), price and cost so far of one server. The root password is left out unless include_password is true (your SSH keys are installed on every server).
입력 스키마
{
"type": "object",
"properties": {
"instance_id": {
"type": "string",
"minLength": 1,
"maxLength": 64,
"description": "Instance id"
},
"include_password": {
"description": "Include the root SSH password",
"type": "boolean"
}
},
"required": [
"instance_id"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🔴create_instance(offer_id, template, disk_gb, name, type, ...)
Rents the server of an offer and starts it with a template. Billing starts right away and runs every minute until you stop or destroy it. Confirm the price with the user first. Needs instances:write. The response has the instance id; poll get_instance until status is running (usually 1–5 minutes).
입력 스키마
{
"type": "object",
"properties": {
"offer_id": {
"type": "string",
"minLength": 1,
"maxLength": 64,
"description": "Offer id from search_offers"
},
"template": {
"type": "string",
"minLength": 1,
"maxLength": 64,
"description": "Template slug from list_templates, e.g. \"comfyui\", \"ollama\", \"pytorch\", \"custom\""
},
"disk_gb": {
"type": "integer",
"minimum": 10,
"maximum": 2000,
"description": "Disk size, GB (use the template's recommended_disk_gb)"
},
"name": {
"description": "A name for the server",
"type": "string",
"maxLength": 60
},
"type": {
"description": "on_demand (default) or interruptible",
"type": "string",
"enum": [
"on_demand",
"interruptible"
]
},
"image": {
"description": "Docker image, only with the custom template",
"type": "string",
"maxLength": 300
},
"ssh_key_ids": {
"description": "SSH key ids to install (default: all of yours)",
"maxItems": 10,
"type": "array",
"items": {
"type": "string",
"maxLength": 64
}
},
"idempotency_key": {
"description": "Reuse the same value when retrying the same request, so a retry never rents twice",
"type": "string",
"maxLength": 64
}
},
"required": [
"offer_id",
"template",
"disk_gb"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}⚪stop_instance(instance_id)
Stops a running server: the GPU is released and billing drops to the disk only. Data on the disk is kept and the server can be started again (the GPU may then be taken by someone else; gpu.nz finds a replacement when it is). Needs instances:write.
입력 스키마
{
"type": "object",
"properties": {
"instance_id": {
"type": "string",
"minLength": 1,
"maxLength": 64,
"description": "Instance id"
}
},
"required": [
"instance_id"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}⚪start_instance(instance_id)
Starts a stopped server again; GPU billing resumes. Confirm with the user. Needs instances:write.
입력 스키마
{
"type": "object",
"properties": {
"instance_id": {
"type": "string",
"minLength": 1,
"maxLength": 64,
"description": "Instance id"
}
},
"required": [
"instance_id"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🔴destroy_instance(instance_id)
Destroys a server and deletes its disk for good; billing stops completely. Irreversible: confirm with the user. Needs instances:write.
입력 스키마
{
"type": "object",
"properties": {
"instance_id": {
"type": "string",
"minLength": 1,
"maxLength": 64,
"description": "Instance id"
}
},
"required": [
"instance_id"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢get_instance_logs(instance_id, tail)
The last lines of a server's container log: boot progress, app start-up, errors. Needs instances:read.
입력 스키마
{
"type": "object",
"properties": {
"instance_id": {
"type": "string",
"minLength": 1,
"maxLength": 64,
"description": "Instance id"
},
"tail": {
"description": "How many last lines (default 200)",
"type": "integer",
"minimum": 1,
"maximum": 1000
}
},
"required": [
"instance_id"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢list_ssh_keys
SSH public keys on the account; they are installed on new servers. Needs user:read.
입력 스키마
{
"type": "object",
"properties": {},
"$schema": "http://json-schema.org/draft-07/schema#"
}🟡add_ssh_key(public_key, name)
Adds an SSH public key (a line from a .pub file) to the account. Needs user:write.
입력 스키마
{
"type": "object",
"properties": {
"public_key": {
"type": "string",
"minLength": 20,
"maxLength": 16000,
"description": "The public key line, e.g. ssh-ed25519 AAAA… you@laptop"
},
"name": {
"description": "A name for the key",
"type": "string",
"maxLength": 60
}
},
"required": [
"public_key"
],
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢topup_link(amount, language)
A link to the balance top-up page (SBP, crypto) for the user to open in a browser; payments never happen through the API. Optionally pre-fills the amount.
입력 스키마
{
"type": "object",
"properties": {
"amount": {
"description": "Amount in the account currency (rubles by default)",
"type": "integer",
"minimum": 1,
"maximum": 10000000
},
"language": {
"description": "Page language (default ru)",
"type": "string",
"enum": [
"ru",
"en"
]
}
},
"$schema": "http://json-schema.org/draft-07/schema#"
}커뮤니티
증거