For developers
One endpoint
An OpenAI-compatible API served by the network. Point your client at it and keep your code. Each response carries a job id you can look up as a receipt.
A permissionless network of graphics cards people already own, serving open models on Robinhood Chain.
Three ways in
For developers
An OpenAI-compatible API served by the network. Point your client at it and keep your code. Each response carries a job id you can look up as a receipt.
For GPU owners
Open the page, pick a model sized to your card, leave the tab running. WebGPU in the browser, or Ollama on CUDA for a full node. Finished jobs are recorded against your wallet.
For holders
The network's token. Launched on Pons and paired with Robinhood's NVDA Stock Token, on Robinhood Chain.
API
The endpoint speaks the chat-completions format, streaming included. The orchestrator hands your request to an idle GPU and relays tokens as they are produced. Prompts and outputs are not stored, only their hashes.
The endpoint opens when the network goes online.
| GPU | Runs as | Model | Tokens / s | Jobs | Wallet | State |
|---|
| Job | Model | Worker | GPU | Tokens | Time | Output hash |
|---|
$NVDAI
Launched on Pons, paired with Robinhood's NVDA Stock Token, traded on Robinhood Chain. Only the contract shown on this page is the real one.