OfficialJev Repos & SDKs
Python SDK Responses
The Python response models: SystemOneResponse groups answers by id, and records the model name, token usage, and request id.
- Format
- Docs
- Published by
- TypeSafe AI
- Last verified
- Oct 2026
Used inJev for Software & DevTools
- Languages
- Python
Highlights
- SystemOneResponse is frozen and strict, and ignores unknown fields. Required fields are model and usage.
- answers is keyed by the question ids you sent. Each value is a Noul, Choice, or Score answer.
- A Choice answer includes the winning label, a 0-1 confidence, and a probability per option.
- usage reports input_tokens and output_tokens. request_id reads the x-typesafe-request-id header.
- raw_http_response exposes the underlying httpx2 response, including status, headers, and body.
Watch out
A generated reference. What confidence means, and when not to act on it, is on the confidence page.
Questions about this entry
Short answers drawn from the linked project or page.
By the question ids you sent, with each value a Noul, Choice, or Score answer. Required fields are model and usage; usage reports input_tokens and output_tokens, and request_id reads the x-typesafe-request-id header.
The winning label, a 0-1 confidence, and a probability per option. SystemOneResponse is frozen and strict and ignores unknown fields, and raw_http_response exposes the underlying httpx2 response.
An expected score that may fall between integer levels, a 0-1 confidence, a legend, and a probability per rubric level. The score is the probability-weighted average of the levels.
A single noul probability from 0 to 1. Values near 1 favor yes or true, values near 0 favor no or false, and values near 0.5 indicate uncertainty.
raw_http_response exposes the underlying httpx2.Response with its status, headers, and body, while request_id reads the x-typesafe-request-id header. The response model itself is frozen and strict.
input_tokens and output_tokens, each an integer or None when the API does not report it. They sit alongside model in the response's required fields.
They sum to approximately 1, with one probability per option in criteria keyed by name. Confidence is separate: it rates the selected choice and can flag uncertain picks for review.
Related terms
Glossary definitions related to this entry.
More Jev resources like this
AsyncTypeSafeClient: async system_one calls, model listing, per-call retry and timeout overrides, and an optional Pydantic response model.
Jev Repos & SDKsDocs#official#docs#sdk
Official
How the Python SDK types questions: Noul, Choice, and Score as objects or dictionaries, including optional JSON instructions and criteria.
Jev Repos & SDKsDocs#official#docs#sdk
Official
Release notes for the official Python SDK, from the 14 September 2026 public release through the Pydantic migration and the HTTP/2 extra.
Jev Repos & SDKsDocs#official#docs#sdk
Official
RetryPolicy for the Python SDK: attempt count, exponential backoff with jitter, which HTTP statuses retry, and a total time budget.
Jev Repos & SDKsDocs#official#docs#sdk
Official
The Python SDK error hierarchy: HTTP status classes, rate-limit wait hints, connection and timeout errors, and response-body validation failures.
Jev Repos & SDKsDocs#official#docs#sdk
Official
Default settings and environment variable names for the Python SDK, including the API root, model alias, and per-request timeout.
Jev Repos & SDKsDocs#official#docs#sdk
Official
From the Jev guides
Original write-ups on what this entry is for.
Closed weights have not stopped the ecosystem from rebuilding the shape in the open. API-compatible servers, retrained heads on small models, and inference-time tricks that skip generation entirely.
Jev Repos & SDKs#open-models#research#local
Read article
From the Jev community
Posts from builders shipping with Jev right now.