> NETWORK_INFRA // RHC_INFERENCE_LAYER

DISTRIBUTED INFERENCE
NETWORK

A mesh of GPU nodes providing AI inference for onchain agents. Each node runs model inference, serves price data, and pushes oracle updates to Robinhood Chain. Run a node, earn rewards.

NETWORK_OPERATIONAL syncing...
ACTIVE NODES
0
connecting...
TOTAL INFERENCE CALLS
0
live counter
AVG LATENCY
--
across all active nodes
NETWORK UPTIME
0s
since last cold start
INFERENCE NODES // LIVE
Querying network...
ARCHITECTURE // REQUEST FLOW
01
User Query
User asks about a stock token via chat or telegram
02
Router Node
Request routed to least-latency inference node
03
Data Fetch
Node reads onchain contract data from RHC RPC + fetches market news
04
Model Inference
LLM processes query with real-time onchain + market context
05
Response / Execution
Answer streamed back. Trading agents execute onchain via non-custodial vault
SUPPORTED MODELS
GLM-4.5 Flash
9B params · 128K context · reasoning
ACTIVE
GLM-5.2
Reasoning model · deep analysis mode
ACTIVE
Custom Node Models
Node runners can plug in their own inference endpoint
ROADMAP
NETWORK ACTIVITY // LIVE
00:00:00[SYS]monitoring started...

RUN A NODE

Provide inference to the network and earn rewards. Minimal requirements: any GPU with 4GB+ VRAM, stable internet, Python 3.10+.

READ DOCS