economy
unified credits system
TOVA Credits are the single unit of account for inference on the network. Instead of paying providers individually, users purchase credits with the TOVA token and spend them across every supported model and provider.
what credits cover
- GPT inference
- Claude inference
- autonomous agents
- coding workflows
- API requests across supported endpoints
- future supported models and providers
request-level transparency
Credits abstract provider pricing while still exposing per-request accounting. Every response includes credits spent, input and output token counts, the selected provider, and the estimated underlying cost where available.
{
"usage": {
"input_tokens": 842,
"output_tokens": 219,
"credits_spent": 0.0184,
"estimated_cost_usd": 0.0123
},
"provider": "groq",
"route": { "objective": "cheapest", "fallback_used": false }
}credit accounting flow
- user purchases credits with the TOVA token
- TOVA tokens used for the purchase are burned on-chain
- credits are issued to the user account
- requests consume credits based on token usage and routing cost
- usage records are surfaced in the dashboard with provider, model, and tx context
Example data: modeled credit spend distribution across workload categories.
info
One balance, every supported model. Funding once and routing dynamically across providers is structurally cheaper than maintaining accounts and quotas with each provider individually.