API · Last updated 2026-09-21

Jev API classify endpoint

A classifier in Jev is a Choice with a closed set (max 255) or a handful of Nouls. There is no documented /v1/classify.

This page is for builders who searched an extra “endpoint” name while wiring Jev / System One. jev.pro is an independent unofficial field guide — not affiliated with TypeSafe AI. We do not sell or issue keys. Confirm first-party routes on docs.typesafe.ai/api.

The public API reference documents one evaluation route — POST https://api.typesafe.ai/v1/systemone — plus GET /v1/models. This title is not a separate first-party URL in that reference. Treat the capability as something you implement around the official call.

Steps

1. Write option descriptions a person could apply. Add other when coverage is incomplete.
2. Mix related Scores (complexity) in the same request.
3. Gate on confidence before you route.
4. Prefer one POST over one-call-per-label HTTP fan-out.

Map onto the official POST

type: "choice" plus a criteria map (unordered). Answer: choice, probabilities, confidence. If you need absolute yes/no per label, use Nouls (noul only, no confidence) instead of stretching Choice. Question ids are not sent to the model.

Pin jev-1.13.0 once you ship a routing table. jev-latest and jev-preview currently resolve to it — confirm. Documented errors: 401, 422, 429, 529. Official SDKs retry 429/529. Vendor-claimed list price meters input only ($0.042 / million; output free). Vendor-claimed 64k / 32k context still applies if you paste a novel into state.

Schema-safe ≠ factually correct. A winner of 0.51 with low confidence is a hold, not a department assignment. Do not copy Noul τ onto Choice confidence. Persist the full probabilities map; the argmax hides a near-tie.

When not to use Choice as “classify”

High-cardinality dumps without descriptions fail the “person could apply this in seconds” test. Open-string labels belong to structured-output LLMs, not System One. JSON-mode classifiers are a different contract.

Pitfalls

• High-cardinality dumps without descriptions.
• Using Choice when you need absolute yes/no per label (use Nouls).
• Calling /v1/classify on api.typesafe.ai.
• Executing rank 2 as a silent second guess.

See also

• API hub — one evaluation POST
• Choice questions — criteria map rules
• Intent routing — classify, then your router
• Route pattern — no /v1/route either

FAQ

Is /v1/classify documented?

No.

255 option cap?

Yes, per official Choice docs.

JSON-mode classifier instead?

Different contract — vs JSON mode.

Does classify return a rationale?

No. Persist the distribution. System One is documented as not writing explanations.

What this page does not claim

• Official TypeSafe status, logos, or support.
• That jev.pro sells, issues, or proxies API keys.
• Extra first-party HTTP routes beyond POST /v1/systemone and GET /v1/models unless the API reference lists them.
• Invented latency SLOs, search volume, or unofficial prices. List price and rate limits are vendor claims — confirm on models.
• That a schema-constrained answer is automatically factually correct.

Disclaimer

This is an independent unofficial site and is not affiliated with TypeSafe AI; official documentation is available at https://docs.typesafe.ai. Never treat jev.pro as TypeSafe official documentation. We do not sell, issue, or proxy API keys.