---
title: "Single best model to use right now (signed routing decision)"
method: GET
path: "/api/premium/route-verdict"
tags: ["Premium"]
---

# Single best model to use right now (signed routing decision)

`GET /api/premium/route-verdict`

The signed model-routing decision. For ?task=code|reasoning|creative|general or ?model=<id-or-name>, returns the single best model to use right now, fusing pricing, contamination-discounted benchmark capability, real production usage, measured p95 latency probes, and live incident-triage operational state, plus runners-up and an AFTA-signed receipt over the inputs. 30-minute operational freshness SLA, no-charge when the live layer is stale.

## Query parameters

- `task` 'code' | 'reasoning' | 'creative' | 'general'
- `model` string
- `max_latency_p95_ms` number
- `require_operational` boolean
- `exclude_deprecated` boolean

## Response `200`

Signed verdict with AFTA receipt.

- object

## Other responses

- `402` — Payment required; returns the x402 challenge.

---

[API](https://skmtc.dev/tensorfeed/apis/tensorfeed-api.md) · [All operations](https://skmtc.dev/tensorfeed/apis/tensorfeed-api/llms.txt) · [OpenAPI document](https://skmtc-service-production.skmtc.workers.dev/v1/apis/tensorfeed/tensorfeed-api/revisions/017e10130065/schema)
