nodeclub.ai
Login
Spots openqwen3.8-27b · 262K

Unlimited tokens.
Flat fee.

Open models on our own GPUs in Europe, with an OpenAI-compatible API. €50/month, without counting a single token.

No credit card · Cancel anytime · Spot reserved for 2 hours

nodeclub · node-01 · Europe · qwen3.8-27bonline
~ $ nodeclub logs --follow
TIMEMODELTOKENSCOST
10:41:15qwen3.8-27b12.418+0,00 €
10:41:20qwen3.8-27b3.902+0,00 €
10:41:27qwen3.8-27b27.551+0,00 €
10:41:36qwen3.8-27b8.840+0,00 €
10:41:40qwen3.8-27b1.736+0,00 €
10:41:45qwen3.8-27b19.204+0,00 €
10:41:52qwen3.8-27b6.512+0,00 €
waiting for requests…
TOKENS THIS MONTH
48.392.117
↑ growing · no limit
COST/TOKEN
0,00 €
YOUR FEE
€50/mo
NODE-01 · SHAREDSPOTS OPEN
Our dedicated GPUs, shared among club members.
OpenAI-compatible API
Infrastructure in the EU
Zero logs
How it works

Change three lines.
Everything else stays the same.

NodeClub speaks the same protocol as the OpenAI API. Your SDK, tools, and code keep working: only the endpoint, key, and model change.

SDKs
Python, Node.js, curl, and any OpenAI-compatible client.
Capabilities
Reasoning, tool calling, and vision.
Context
262K tokens: entire repos and long documents.
app.py+3−3
1from openai import OpenAI
2
3client = OpenAI(
4−base_url="https://api.openai.com/v1",
4+base_url=NODECLUB_URL,
5−api_key=OPENAI_API_KEY,
5+api_key=NODECLUB_KEY,
6)
7
8resp = client.chat.completions.create(
9−model="gpt-4o",
9+model="qwen3.8-27b",
10messages=messages,
11)
Membership

All inclusive.
Nothing to watch over.

Flat rate, flat invoice
With per-token APIs, every test adds up and every month varies. Here you pay the same every month.
NodeClub · €50/moPer tokenmayjunjulagosepoct
Zero logs, in the EU
We do not store the content of your requests.
Public status
Real-time cluster status, load, and speed.
Your API key
A personal key for all your tools, and your usage dashboard.
nc-••••••••••••3f2a
Your usage · 7 daysNo limit
Community votes the next model
Members decide what model enters the cluster next.
Which model should we add next?
Gemma
Next Qwen
Propose one in #votaciones
Community

A club,
not a vendor.

Behind every API key is a developer. In the private member Discord, get help, build agents, and vote on upcoming models.

nc
NodeClub
CHANNELS
#general
#agentes
#anuncios
#us-members
#ayuda
#votaciones
#ayudaQuestions among members
LR
Lucía R.today at 10:42
Anyone using the 262K context for RAG over an entire codebase repo? How does it perform?
DM
Dani M.today at 10:44
I am running it with OpenCode on a large monorepo. Since the latest cluster node update, resuming long context sessions is virtually instant.
73
LR
Lucía R.today at 10:46
Awesome, setting it up this afternoon.
MS
Marta S.STAFFtoday at 10:48
Deploying cluster improvements tonight. We will post details in #announcements before anything changes.
9
Send message to #ayuda
Model

What runs today,
no fine print.

We publish real limits for every model: concurrency, speed, context, and service status. Upcoming models are chosen by the community.

qwen3.8-27bNVFP4Active
Tokens per month
Unlimited
Requests per minute
30
Parallel requests
3
Context window
262K
Capabilities
Reasoning · Tools · Vision
Pricing

One price. All included.

MemberSpots open
€50/mo

No commitment. Cancel anytime.

24h trial, no card needed

Includes
  • Unlimited monthly tokens
  • qwen3.8-27b with 262K context
  • 30 requests/min, 3 parallel
  • Your personal API key and dashboard
  • Private member Discord
  • Vote on next models
  • Zero logs, processed in the EU
Enterprise
Dedicated node

For fine-tunes, LoRAs, or custom models outside the shared cluster. Managed by our team with a fixed monthly fee.

Contact us
FAQ

Frequently
asked questions.

Have questions? Contact us.

How does the 24-hour trial work?

When spots are open, request access with your email and we reserve your spot for 2 hours. Entering via the link activates 24 hours of full access, no card required. If you don't join afterwards, access simply expires.

Is there really no token limit?

There is no monthly token cap. We do enforce rate limits to ensure healthy performance for all members: 30 requests per minute and 3 concurrent requests per member, under a fair-use policy.

Do you log prompts or completions?

No. Zero logs: we do not store the content of your requests, and everything is processed inside the European Union.

What exactly is the member Discord?

A private Discord server for members. Get help directly from other developers, showcase projects, vote on upcoming models, and get advance notices of cluster updates.

Is there a long-term commitment?

No. You pay €50/month and can cancel anytime.

What if my company needs a custom model?

For fine-tunes or custom models, we don't use the shared cluster: we deploy a fully managed dedicated node at a fixed monthly rate.

Test the cluster.
24 hours, free.

No card required. Spot reserved for 2 hours once requested.