API · Last updated 2026-09-21

Jev API evaluate endpoint

SDK methods are named system_one / systemOne. Some Vercel AI SDK writeups use experimental_evaluate against the Gateway — that is a Gateway/SDK adapter, not a second TypeSafe path we will invent.

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. First-party: POST /v1/systemone or official SDK systemOne.
2. Gateway: follow current Vercel docs for method names and model id typesafe-ai/jev.
3. Do not call /v1/evaluate unless TypeSafe lists it.
4. Log the resolved model.

Naming vs contract

Same body: state, model, questions. Same answers map. Choice (max 255) and Score (2–10) return probabilities + confidence. Noul returns noul only. Question ids are not sent to the model. Text only; English primary.

Python ≥ 3.10: typesafe-sdk, TypeSafeClient / AsyncTypeSafeClient, TYPESAFE_API_KEY. Node 20+: @typesafe-ai/sdk, TypeSafeClient.systemOne. Official SDKs retry 429/529. Other documented statuses: 401, 422.

jev-latest and jev-preview currently resolve to jev-1.13.0 — confirm. The Gateway spelling typesafe-ai/jev is not the first-party model field. Mixing Gateway method names with the first-party host is the usual 401. “Evaluate” does not return a rationale; System One is documented as not writing explanations.

jev.pro does not expose evaluate and will not proxy a call. Vendor claims (list price $0.042 / million input, output free; 250k tok/s; 1200 RPM; 70ms–500ms launch-post range) are not SLOs this page measured.

When the word “evaluate” is fine

Use it in your own service names. Do not put it on api.typesafe.ai as a path. LangChain TypeSafeClassifier.invoke is a LangChain wrapper, not TypeSafe.

Pitfalls

• Mixing Gateway method names with the first-party host.
• Assuming “evaluate” returns a rationale.
• Calling /v1/evaluate.
• Sending typesafe-ai/jev to api.typesafe.ai.

See also

• API hub — first-party paths
• HTTP endpoint — POST and GET aliases
• First API call — curl
• Python SDK — system_one

FAQ

Is /v1/evaluate documented?

Not in the reference we used. The path is /v1/systemone.

LangChain TypeSafeClassifier.invoke?

A LangChain wrapper — harness.

Does jev.pro expose evaluate?

No. We do not proxy.

Why do writeups say evaluate?

Adapter method names on other doors. Confirm that door’s docs.

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.