TVP API

Build against live model pricing and routed AI inference.

TVP keys use an OpenAI-compatible surface for model discovery, chat completions, and embeddings while keeping usage and key status tracked in the TVP registry.

Get API Key Manage Keys

GET/v1/models

List live models

Returns the synced live provider catalog with model IDs, names, context, modalities, pricing, and metadata.

curl https://theverificationplatform.com/v1/models \
  -H "Authorization: Bearer tvp_live_YOUR_KEY"

POST/v1/chat/completions

Run chat completions

Forwards OpenAI-compatible chat requests through the TVP gateway and tracks usage against the TVP key.

curl https://theverificationplatform.com/v1/chat/completions \
  -H "Authorization: Bearer tvp_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"openrouter/owl-alpha","messages":[{"role":"user","content":"Say hello from TVP"}]}'

POST/v1/embeddings

Generate embeddings

Forwards OpenAI-compatible embedding requests for embedding-capable models when the upstream provider supports the model.

curl https://theverificationplatform.com/v1/embeddings \
  -H "Authorization: Bearer tvp_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"qwen/qwen3-embedding-8b","input":"TVP model procurement"}'

Limits and model support

Key status enforcement

Revoked or paused keys are blocked before any upstream provider call is made.

Request limits

Each key has a monthly request cap and a short-window rate limit enforced server-side.

Provider availability

Final model availability depends on upstream endpoint health, provider permissions, and supported modalities.

Usage tracking

TVP records endpoint, status, model, and token usage when returned by the provider response.