API · Last updated 2026-09-21

Jev API validate endpoint

TypeSafe already validates the evaluation body (422). Semantic validation (“does this invoice look complete?”) is a Noul or Score you write. There is no documented /v1/validate.

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. Fix 422s by reading the official field error.
2. Put completeness / consistency questions in questions.
3. Remember schema-safe ≠ factually correct.
4. Keep regex and arithmetic in code (extraction cookbooks).

Two different “valid”

HTTP 422 means model, state, or questions failed the documented shape (missing field, bad type, Choice over 255, Score outside 2–10). Official SDKs do not retry that. 200 means the request parsed — not that the invoice is real.

Semantic checks are ordinary primitives on the same POST: Noul for yes/no completeness, Score for ordered quality, Choice for “which required section is missing.” Noul has no confidence field; gate on the probability. Choice/Score expose confidence. Question ids are not sent to the model.

A dry-run flag we will invent does not exist. Your own JSON Schema can cheaply reject bodies before you spend input tokens (vendor claim: $0.042 / million; output free). Pin jev-1.13.0 so semantic fixtures do not drift with jev-latest (currently the same id as jev-preview — confirm). Other documented errors: 401, 429, 529 (SDKs retry the last two).

When not to use Jev as a validator

Do not use Score expectations as checksums. Do not ask Jev to verify HMAC signatures or arithmetic totals. Structured-output LLMs that fill JSON are a different product — compare vs structured outputs.

Pitfalls

• A dry-run flag we will invent.
• Using Score expectations as checksums.
• Calling /v1/validate.
• Treating 200 as “the document is genuine.”

See also

• API hub — body contract
• Request format — required fields
• Schema mismatch — 422 symptoms
• Schema validation — term

FAQ

Is /v1/validate documented?

No.

Can I validate without spending tokens?

Only your own JSON schema. Semantic Jev questions cost input tokens (vendor pricing).

Structured outputs vs this?

vs structured outputs.

Does 422 include a usage block?

Do not assume it. Meter successful calls.

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.