Network

Independent providers become one logical AI infrastructure layer.

The network connects users, AI agents, local models, independent compute providers, decentralized marketplaces and temporary clusters coordinated by SOLAI Coder.

Users

Run AI locally, rent remote compute, or authorize an agent to spend within explicit compute limits.

Providers

Contribute idle CPU/GPU capacity, set prices, define schedules, advertise compatible models and retain hardware control.

Agents

Search the marketplace, select providers, create clusters and complete complex work under user-defined budgets.

Two Marketplaces

Buy compute directly or buy an outcome through a cluster.

Level 1 is provider rental. Level 2 is SOLAI Coder assembling providers into a temporary execution environment for a specific result.

Compute Marketplace

Filter providers by model, GPU, VRAM, price, latency, uptime, reputation, region and availability.

solai compute list --model SOLAI-80B --max-price 15

Cluster Marketplace

Submit a task and let SOLAI Coder determine the worker count, model mix, deadline strategy and maximum budget.

solai coder "Analyze 50,000 documents"
Provider Lifecycle

Voluntary participation with explicit boundaries.

SOLAI is an inference network, not a remote shell marketplace. Consumers submit AI jobs; providers define resource limits and can leave under graceful shutdown rules.

1

Register

Wallet, hardware and model metadata.

2

Heartbeat

Signed status and capacity reports.

3

Available

Provider appears in discovery index.

4

Busy

Jobs execute through runtime limits.

5

Settle

Validated work receives escrow payment.