Build without
limits.

Don't run out of tokens. Ever.

NaN is the community of builders sharing GPUs to run AI open models. Build whatever you need, without paying per token. You can just do things.

01what is nan

What is NaN.
The Builders Community

Builders sharing dedicated GPUs to run open models, with no bill per token.

Closed-model APIs are convenient but expensive and limited. Open models are free but running them seriously requires hardware almost no one has at home.

NaN solves this by sharing powerful GPUs among builders who actually use them, splitting the real cost of infrastructure.

Open to anyone who wants to build. Creating an account is free; the cluster, the community and the workspaces switch on with a plan.

The community lives on Discord. The website only manages access. Every member pays to be here. Only those who truly want to build make it in.

02desktop app

Helmcode Desktop
for desktop

NaN's models in a native app on your machine. Download it, sign in and start building.

Requirements: macOS 12+, Windows 10+, Ubuntu 22.04+ or equivalent. All downloads and install notes →

03built with nan

We don't burn tokens.
We build.

Real projects built by NaN members using the shared inference cluster. This is what happens when builders get serious hardware.

orca-nan

@rub3n88

beyond-the-protocol
NaN·MF2-HD1.44 MB Beyond The Protocol BeyondTheProtocol es una webapp médica privada construída para una paciente con cáncer metastásico ultra-raro. Problema que resuelve: gestionar cientos de documentos clínicos dispersos (PDFs, análisis, reportes) de forma inteligente. Centraliza todo en un sistema RAG bilingüe (ES/EN) con búsqueda semántica, dashboard longitudinal, y un wiki clínico compilado, permitiendo explorar el historial médico completo con contexto y continuidad. websites · ai

@danitoszwarc

openbide

@rokkoo

kit72h

@ntizar

resomer

@angelfg

forja-de-historias

@hensip

https-www-tokensanonimos-com

@rokkoo

agentbrowser

@aitor

spendscopio-ai-expense-tracker

@sergi_ruiz

bender

@barckcode

nan-credits-for-orca

@javierorp

nan-credits-for-gnome

@javierorp

fuel-radar

@javierorp

precio-luz-pvpc

@javierorp

aimigo

@javierorp

ruben-martinez-interactive-developer-portfolio

@rmartbernabe

04the cluster

19 models.
One API.

The full open-model inference stack, voted by the community every quarter. Data stays in the EU, with nothing logged.

// llm
deepseek-v4-flash most used frontier
DeepSeek

305B MoE · FP8 · 1M context · vision · tool calling · reasoning

credits
mimo-v2.6-flash frontier
Xiaomi

omnimodal · 1M context · vision · audio · tool calling · reasoning

credits
hy4-flash frontier
Tencent

1M context · vision · tool calling · reasoning

credits
glm5.3 frontier
Z.ai

753B MoE · NVFP4 · sparse attention · 1M context · text only · tool calling · reasoning

credits · larger model
qwen3.8-flash frontier
Alibaba

125B-6B MoE · 1M context · vision · tool calling · reasoning

credits
glm5.3-flash frontier
Z.ai

320B-18B MoE · FP8 · 1M context · multimodal · tool calling · reasoning

credits
kimi-k3.1-pro frontier
Moonshot AI

1M context · vision · tool calling · reasoning

credits · larger model
qwen3.8-max frontier
Alibaba

1M context · vision · tool calling · reasoning

credits · larger model
qwen4-max frontier
Alibaba

1M context · vision · tool calling · reasoning

credits · larger model
deepseek-v4-pro frontier
DeepSeek

1M context · tool calling · reasoning

credits · larger model
hy4 frontier
Tencent

1M context · vision · tool calling · reasoning

credits · larger model
gemma4
Google

26B-A4B MoE · FP8 · 262K context · tool calling · reasoning · vision

unmetered
qwen3.6 most used
Alibaba

35B-A3B MoE · FP8 · 262K context · tool calling · reasoning · vision

unmetered
// embedding
qwen3-embedding
Alibaba

8B · 4096 dims · 100+ languages · MMTEB 70.58

unmetered
// rerank
rerank
Alibaba

Qwen3-Reranker-8B · BF16 · 100+ languages · cross-lingual

unmetered
// tts / stt
kokoro
Hexgrad

TTS · 82M params · 67 voices · <1s latency · ef_dora · em_alex

unmetered
whisper
OpenAI

STT · large-v3 · CTranslate2 · ~3.2% WER Spanish · auto language detection

unmetered
// image
flux-2-klein
Black Forest Labs

FLUX diffusion · text→image · image→image · 256-1536 px · 1-4 per request

100 req / mo
qwen-image-2.1
Alibaba

Qwen diffusion · text→image · 512-1280 px · 1-4 per request

100 req / mo
05privacy

Pure inference.
Zero logs.

NaN is inference infrastructure. There's no logging of prompts, model responses, nothing. Everything lives in your local session, on whatever client, agent or custom app you use.

  • Zero logs. No prompt or response logging.
  • Everything lives in your local client or agent session.
  • Processing in the European Union.
  • Your code doesn't train any models.
  • Server metrics only, for cluster maintenance (tokens/s, RPM).
06membership

One community.
Two tiers. One premium path.

The community is the product. Inference is the superpower you add on nan_member; premium brings the most credits and concurrency for members.

nan_member · glm 5.3 premium glm 5.3 · premium
200€ /mo · VAT included
  • 20,000 credits every billing period for every chat model, GLM 5.3 and the larger ones included
  • Credits come back in full when your billing period starts; credit packs 5% cheaper if you need more
  • 1M context · 10 concurrent requests
  • One free Micro workspace (1 vCPU · 2 GiB RAM · 10 GiB disk)
  • Everything in nan_member
  • Access switches on a few minutes after payment
Go premium

Month to month, no commitment. Billed in euros from any region.

nan_member inference included
70€ /mo · VAT included
  • Access to the shared cluster. The models we host ourselves draw no credits.
  • 7,000 credits every billing period for every chat model, the larger ones included
  • Open models chosen by the community
  • Personal API key compatible with OpenAI
  • Private Discord channels for members only
  • Events, workshops and hackathons
  • Voice in the quarterly model vote
Get inference

Month to month, no commitment. Billed in euros from any region.

nan_community
14,99€ /mo · taxes included
  • Discord access
  • Member-only channels
  • Live discussions, builds, AMAs
  • Events, workshops and hackathons

Discord and events only: no credits and no API key included. To call the models, buy credit packs or get nan_member (70€/mo), which includes 7,000 credits every billing period.

Join the community

No commitment. Cancel anytime. Billed in euros from any region.

workspaces · add-on

Your own cloud machine.

A dedicated, isolated dev box reachable over SSH and from the browser, with Pi, Hermes, gentle-shell and Herdr ready to use. Anyone with a NaN account can buy slots; each slot keeps one workspace running.

size vCPU · RAM · disk slot daily backups
micro 1 · 2 GiB · 10 GiB 4,99€/mo +0,99€/mo
nano 2 · 4 GiB · 20 GiB 8,99€/mo +1,49€/mo
basic 4 · 8 GiB · 40 GiB 16,99€/mo +2,49€/mo
medium 8 · 16 GiB · 80 GiB 32,99€/mo +3,99€/mo
large 8 · 32 GiB · 160 GiB 60,99€/mo +6,99€/mo
  • The premium plan includes one free Micro workspace.
  • Slots pay for the machine only. Agents use NaN inference only if your plan includes it.
  • Backups are optional: one snapshot a day, kept 7 days, restore it yourself.
  • Monthly, billed in euros. Cancel a slot anytime.
07who's behind

The NaN community is led by Cristian Córdova and Borja Perez.

I started NaN because I wanted to run serious open models without usage limits, and share and learn with people in the same boat: builders who actually burn tokens.

Cristian Córdova
Cristian Córdova

Cristian Córdova

@barckcode · Tenerife, Islas Canarias

Founder of Helmcode

Founder of Helmcode, working with startups on cloud infrastructure, AI tooling and open source.

Thanks to Helmcode we offer open model inference with unlimited token plans for enterprises. I share my experiments on social media and AI research.

Borja Perez

Borja Perez

@borjaperfra · Valencia, Spain

Marketing Lead

Builder with marketing skills. Helping Cristian with comms and community.

Our goal is to create the best open-models community in the world, helping people create and ship products with open AI.

08faq

What you'll probably ask.

What is NaN?

A closed community of builders sharing dedicated GPUs to run open models. Every member pays to be here, and there's only room for people who actually build.

How does the cluster work?

The models run 24/7 on a shared inference cluster. Each nan_member (70€/mo, inference included) gets a personal API key compatible with OpenAI: change the base_url and you're in. The 14,99€ NaN Community plan runs no models and has no API key.

Is it really unlimited tokens?

On the cluster models, no counter. The rest are paid from credits so the plan holds up: nan_member includes 7,000 credits per billing period and premium 20,000, one pool shared by every model. Each request draws credits by the model's price per token, so the flash models go much further than the larger ones. The price of each model is on its card in the portal, not in fine print. The bill is the fixed subscription price and it doesn’t move, unless you choose to buy credit packs.

How are the models chosen?

Every quarter members vote on Discord which new models join the cluster. Candidates are pre-selected technically: they have to fit the real capacity. The open ecosystem moves fast, and so does the cluster.

Where does my data go?

Processed in the European Union, with no logging of prompts or responses. Your API key is personal and non-transferable. Your code trains no model and your data never leaves your session.

Who is NaN for?

Builders who want to run open models seriously without building the infrastructure: indie hackers, makers, people with agents and side projects in production. It's not for someone who just wants to try a couple of prompts.

Why only open models?

Because they can be audited, modified and used without asking permission. Closed ones are convenient but create dependency. With open models, the stack is yours.

Can I train models?

No. NaN is inference, not training. The GPUs are optimized to run existing models at high speed, not for fine-tuning.

What is a workspace?

Your own cloud machine for development: a dedicated, isolated box you reach over SSH or from the browser, with coding agents such as Pi and Hermes ready to use. Anyone with a NaN account can buy workspace slots from 4,99€ a month, and the premium plan includes one free Micro workspace. A slot does not include inference: agents use NaN models only on an inference plan. The workspaces guide in the docs has sizes, prices, network rules and how backups work.

What payment methods do you accept?

Credit or debit card. New memberships are billed in euros from any region, taxes included; subscriptions opened earlier in US dollars keep their original price. No commitment: cancel anytime.

How do I get in?

Create your account with your email: it's free and takes a minute. Then pick what you need (community, inference, premium or a workspace) and it switches on right after payment.