Models to serveRTX 5090 · 32 GB
qwen3.8-omni-flash
$0.03/M
deepseek-v4.1-flash
$0.02/M
kimi-k3
$0.05/M
Set your own price.
You choose what to serve and the terms on which you offer it.
Your own provider
Offer open models on your GPUs, capacity you hold, a router you built, a specialized service or an agent.


Active users
35,501Requests served
35.8MUSDC settled
$419.3KProviders
304Running your provider
Choose what you serve and at what price. Requests that match your terms are routed to you and served automatically.
Models to serveRTX 5090 · 32 GB
qwen3.8-omni-flash
$0.03/M
deepseek-v4.1-flash
$0.02/M
kimi-k3
$0.05/M
You choose what to serve and the terms on which you offer it.
Incoming requestServed
deepseek-v4.1-flash · 860 tokens
Incoming requestnow
deepseek-v4.1-flash· 2,140 tokens
from 0x5e…a2 · at your $0.02/M
Serving… Served · earned $0.000043
Requests that match your models and prices arrive and are served automatically.
Your reputation
Settled usage and the ANTS staked behind your provider give buyers a reputation they can check. Responses are signed, and model fingerprints help verify what you serve.
Foundation Provider Staking Program
The Foundation may stake ANTS behind active, high-performing providers. Support depends on agreed milestones and requires KYB/KYC screening. Providers that use a TEE or offer a free model receive priority.
Foundation staking
Antseed runs on your computer. Your chat, your agents and your apps all go through it.