Node 01 is online

AI that runs
on sunlight.

SolarNodes serves open models from a micro data center at a remote location. Powered by a solar array, buffered by batteries, cooled by passive air flow. No grid. No fossil fuels. Every answer comes straight from daylight.

Try Node 01 See how it works

The path of a request

Sunlight in. Tokens out.

See the full pipeline →

Solar array

Panels on the container roof convert daylight to DC power.

Battery bank

Surplus fills the bank for clouds, dusk, and overnight runs.

Inference cluster

Open models load on local hardware inside the container.

Your request

Chat or API.

Available models

Open weights on Node 01.

These models run entirely on our own hardware using publicly released weights. Your requests never reach Meta, Alibaba, or any other model maker's servers — no API fees, no usage data, and no revenue flows back to them. How open-weight models work

Llama 3.1

Meta's instruction-tuned open model — the fast default on Node 01 for real-time chat when latency and solar budget matter most.

llama.meta.com

Qwen 2.5 14B

Alibaba's multilingual open model with strong instruction following across dozens of languages and solid coding ability — higher quality, higher latency.

qwen.ai

Why solar

The data centers powering today's AI strain grids, drain water, and burden our communities.

While data centers are the invisible backbone of our digital world, their rapid expansion comes with a heavy environmental and local toll. Carbon emissions remain a primary concern, as these massive facilities consume vast amounts of electricity 24/7, often straining local power grids and relying on fossil fuels to maintain uptime. Beyond their carbon footprint, their immense cooling needs lead to severe resource strain and water contamination; data centers dissipate heat by evaporating millions of gallons of water daily, sometimes discharging chemically treated wastewater back into local ecosystems or depleting precious municipal supplies.

Furthermore, surrounding communities are increasingly battling infrasound and low-frequency noise pollution generated by massive industrial cooling fans and backup diesel generators. This constant, low-frequency hum can travel for miles, leading to sleep deprivation, anxiety, and a diminished quality of life for nearby residents, turning these digital hubs into material environmental hazards.

The cost of
Hyperscaling AI infrastructure.

EESI

Data centers and water consumption

Large facilities can withdraw millions of gallons per day for cooling, drawing down freshwater supplies in communities already facing drought and shortages.

Read article

EESI

Communities raising noise concerns

Residents near data centers report constant humming from cooling systems and diesel generators, with little recourse under local noise ordinances.

Read article

SELC

Illegal on-site power plants for AI

xAI installed dozens of unpermitted gas turbines to power a data center, releasing smog-forming pollution near homes, schools, and churches.

Read article

API access