← All commodities
◆

Compute & Inference

Rent raw GPU/CPU compute or ready-made AI inference. Pay per token, per second, or per subnet.

Agent quick-start (view → connect → pay → use)

  1. View price: GET /api/v1/providers?category=compute (filter by price/unit)
  2. Connect: call the provider endpoint or SDK listed per provider
  3. Pay: prepaid credit, API key, or on-chain (subnet) as specified
  4. Use: send workload; copy the minimal snippet into your agent loop

Providers

OpenAI API

live
Price$0.002 / 1k input tokens
Unittokens
Min$5
Connecthttps://api.openai.com/v1
Authapi_key
Payprepaid
RateTPS-based
import openai
r=openai.chat.completions.create(model="gpt-4o",messages=[{"role":"user","content":"hi"}])
source →

Together AI

live
Price$0.0008 / 1k tokens (Llama-3-70B)
Unittokens
Min$0
Connecthttps://api.together.xyz/v1
Authapi_key
Payprepaid
requests.post("https://api.together.xyz/v1/chat/completions",headers={"Authorization":"Bearer KEY"})
source →

Bittensor SN21 (Gradients)

live
Pricealpha-locked
Unitsubnet
Minstake
Connectsubnet 21 REST
Authwallet
Payonchain
# query subnet 21 via bittensor; lock alpha for access
source →