Live on Robinhood Chain

The open-source
AI compute layer
for everyone.

GrossFi is a decentralized inference network powering open-source language models. Run open-source AI workloads on distributed hardware. Earn rewards. Build the future of intelligence.

Run a Node Soon How It Works
Scroll
0
Active Nodes
0
Models Hosted
0
Inference/sec (K)
0
Network Uptime
Infrastructure

Built for open-source AI

GrossFi provides the decentralized compute backbone for running open-source language models at scale. No gatekeepers. No API keys. Just distributed intelligence.

◆

Distributed Inference

Workloads are sharded across node operators globally. Low latency, high throughput, zero single points of failure.

Sharded
◈

Open-Model Runtime

Optimized for open-source model families — full GGUF/ONNX support with hardware-aware scheduling across GPU and CPU nodes.

GGUF / ONNX
◉

Stake-to-Compute

Node operators stake $GROSS to participate. Higher stake = higher task priority. Slashing ensures honest computation.

Slashing
◊

Proof of Inference

Cryptographic verification of compute work. Every inference job is attested on-chain. Fraud proofs protect requesters.

On-chain attested
○

Robinhood-Native Settlement

All rewards, payments, and staking settle on Robinhood Chain. Sub-cent gas fees. Finality in seconds.

Robinhood Chain
◌

Open Model Registry

Anyone can register open-source models. Community-governed curation ensures quality. No centralized model gatekeeping.

Community-governed
Network

Global node distribution

GrossFi operators span 40+ countries. The network self-balances workloads based on latency, capacity, and hardware capability.

42 Regions Online
1,347 Nodes Active
Process

How GrossFi works

Four steps to participate in the decentralized inference economy.

1

Stake $GROSS

Deposit tokens to activate your node. Minimum stake scales with hardware tier. Higher stake unlocks priority task routing.

2

Receive Tasks

The network scheduler routes inference jobs to your node based on hardware capability, latency, and stake tier.

3

Run Inference

Execute workloads using local GPU/CPU resources. Submit cryptographic proofs of computation for each completed task.

4

Earn Rewards

Verified compute earns $GROSS rewards per task. Rewards scale with work completed, uptime, and hardware contribution.

Token

$GROSS tokenomics

100B supply. Fair launch on Bankr. 0% tax. 0.7% swap fee funds the protocol.

100B
Total Supply
Fair Launch
Bonding curve on Bankr — unlocked at TGE
85%
Creator Allocation
30-day cliff, 2-year linear vesting
15%
Timeline

Roadmap

Phase 1 — Q2 2026

Genesis Launch

Token launch on Robinhood Chain. Initial node operator onboarding. Core inference engine deployment with open-source model support.

Token Deploy Node Software v1 Testnet Live
Phase 2 — Q3 2026

Network Expansion

Mainnet node activation. Multi-model registry. Staking contracts. First enterprise inference partnerships.

Mainnet Nodes Staking Live Model Registry
Phase 3 — Q4 2026

Proof of Inference

On-chain compute verification. Fraud proof system. Cross-chain inference requests from Ethereum, Optimism, Arbitrum.

PoI Protocol Cross-Chain Governance
Phase 4 — Q1 2027

Autonomous Intelligence

DAO governance activation. Autonomous model fine-tuning pipelines. Agent-to-agent inference marketplace.

DAO Launch Agent Market Fine-Tune Pipelines
Participate

Node operator tiers

Choose your hardware tier. Higher tiers receive priority routing and increased reward multipliers.

★

Edge Node

Consumer GPU (8GB+ VRAM). Run quantized models for lightweight inference tasks. Perfect for community contributors.

Min Stake: 11,200,000 $GROSS · Reward: 1x
★★

Compute Node

Professional GPU (24GB+ VRAM). Full-precision inference. Priority task routing. Serve enterprise-grade workloads.

Min Stake: 22,200,000 $GROSS · Reward: 2.5x
★★★

Validator Node

Multi-GPU cluster. Run inference + validate proofs. Highest reward tier. Participate in network governance.

Min Stake: 44,200,000 $GROSS · Reward: 5x
♦

Model Curator

Stake without hardware. Curate the model registry. Vote on model quality. Earn curation fees from inference volume.

Min Stake: 84,200,000 $GROSS · Reward: Variable
Questions

FAQ

What hardware do I need to run a node?+
Minimum: any GPU with 8GB VRAM (RTX 3060 or equivalent) or 16GB RAM for CPU-only nodes. Recommended: RTX 4090 or A100 for maximum rewards. The network auto-schedules tasks based on your hardware capability.
How are inference rewards calculated?+
Rewards = (tasks_completed × task_weight × uptime_multiplier × tier_multiplier). Tasks are weighted by computational complexity. Uptime above 95% gives bonus multipliers. Higher staking tiers unlock higher tier multipliers.
Is GrossFi affiliated with Nous Research?+
GrossFi is an independent community project inspired by the open-source AI movement. We are built on the same principles: open models, decentralized access, and community-driven development. Our inference engine is optimized for the open-source model families including Llama, Mistral, and Qwen.
What models can I run on GrossFi?+
Any open-source model registered in the Model Registry. Currently optimized for: Llama 3, Mistral, Mixtral, Phi-3, Qwen, DeepSeek, and other open-source architectures. GGUF and ONNX formats supported. Community can register new models through governance proposals.
How does Proof of Inference work?+
Each inference task includes a cryptographic challenge. Nodes must produce output that matches the expected result within tolerance. Random spot-checks are performed by validator nodes. Fraudulent results trigger slashing of the operator's staked $GROSS.
When token launch?+
$GROSS token deployment on Robinhood Chain is targeted for Q2 2026. Follow our social channels for exact date announcement. Early node operators receive Genesis NFT badges with permanent reward multipliers.

The future of AI is
decentralized.

Become an early node operator. Earn Genesis rewards. Help build the open compute layer for the next generation of intelligence.

Launch Node Operator Soon