HackerReposA FIELD GUIDE TO OPEN SOURCE
ProSponsor the shelf♧ 0
← Back to top all-time
fstandhartinger
jevbench
Python2026–2026
fstandhartinger / jevbench

jevbench

JevBench by Benchmark Heaven: benchmark and leaderboard for AI decision models, measuring typed decision accuracy, calibration, latency and cost.

3Distinct people
4Mentions
0Explicit recommenders
341GitHub stars
REPOSITORY AT A GLANCE

What to know before exploring

What it is

JevBench by Benchmark Heaven: benchmark and leaderboard for AI decision models, measuring typed decision accuracy, calibration, latency and cost.

GitHub

Where it is used

No documented product use has been verified yet.

What people are saying

No recommendation phrase was found in the indexed comments.

No classified warning is attached to this page. Read the discussions before treating silence as approval.

Also in these discussions

No other repository is repeatedly linked in the same discussions.

This week on Hacker News

Nov 9–Nov 16 UTC · Direct repository links and separately labeled related project stories. Indexed through 11/16/2026, 8:59:02 AM UTC.

No direct repository links were found in the indexed HN discussions over the past seven days. Older discussions are available below.

Mentions over time

HN mentions by hour4 dated mentions

Observed mentions: 2026-09-22 13:01:03 UTC to 2026-09-23 11:32:15 UTC.

012026-09-22 13:00:00 UTC to 2026-09-22 14:00:00 UTC (exclusive): 1 mentions; 1 submissions, 0 commentsSep 22, 13:002026-09-22 14:00:00 UTC to 2026-09-22 15:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-22 15:00:00 UTC to 2026-09-22 16:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-22 16:00:00 UTC to 2026-09-22 17:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-22 17:00:00 UTC to 2026-09-22 18:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-22 18:00:00 UTC to 2026-09-22 19:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-22 19:00:00 UTC to 2026-09-22 20:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-22 20:00:00 UTC to 2026-09-22 21:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-22 21:00:00 UTC to 2026-09-22 22:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-22 22:00:00 UTC to 2026-09-22 23:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-22 23:00:00 UTC to 2026-09-23 00:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-23 00:00:00 UTC to 2026-09-23 01:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 commentsSep 23, 00:002026-09-23 01:00:00 UTC to 2026-09-23 02:00:00 UTC (exclusive): 1 mentions; 0 submissions, 1 comments2026-09-23 02:00:00 UTC to 2026-09-23 03:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-23 03:00:00 UTC to 2026-09-23 04:00:00 UTC (exclusive): 1 mentions; 0 submissions, 1 comments2026-09-23 04:00:00 UTC to 2026-09-23 05:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-23 05:00:00 UTC to 2026-09-23 06:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-23 06:00:00 UTC to 2026-09-23 07:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-23 07:00:00 UTC to 2026-09-23 08:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-23 08:00:00 UTC to 2026-09-23 09:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-23 09:00:00 UTC to 2026-09-23 10:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-23 10:00:00 UTC to 2026-09-23 11:00:00 UTC (exclusive): 0 mentions; 0 submissions, 0 comments2026-09-23 11:00:00 UTC to 2026-09-23 12:00:00 UTC (exclusive): 1 mentions; 0 submissions, 1 commentsSep 23, 11:00

Counts come from the repository’s dated source evidence. Index cutoff: 2026-11-02 08:16:59 UTC.

View mention counts and date ranges
UTC ranges include their start and exclude their end. Select a period to filter the source evidence.
PeriodStart (UTC)End (exclusive, UTC)MentionsSubmissionsComments
Sep 22, 13:002026-09-22 13:00:00 UTC2026-09-22 14:00:00 UTC110
Sep 22, 14:002026-09-22 14:00:00 UTC2026-09-22 15:00:00 UTC000
Sep 22, 15:002026-09-22 15:00:00 UTC2026-09-22 16:00:00 UTC000
Sep 22, 16:002026-09-22 16:00:00 UTC2026-09-22 17:00:00 UTC000
Sep 22, 17:002026-09-22 17:00:00 UTC2026-09-22 18:00:00 UTC000
Sep 22, 18:002026-09-22 18:00:00 UTC2026-09-22 19:00:00 UTC000
Sep 22, 19:002026-09-22 19:00:00 UTC2026-09-22 20:00:00 UTC000
Sep 22, 20:002026-09-22 20:00:00 UTC2026-09-22 21:00:00 UTC000
Sep 22, 21:002026-09-22 21:00:00 UTC2026-09-22 22:00:00 UTC000
Sep 22, 22:002026-09-22 22:00:00 UTC2026-09-22 23:00:00 UTC000
Sep 22, 23:002026-09-22 23:00:00 UTC2026-09-23 00:00:00 UTC000
Sep 23, 00:002026-09-23 00:00:00 UTC2026-09-23 01:00:00 UTC000
Sep 23, 01:002026-09-23 01:00:00 UTC2026-09-23 02:00:00 UTC101
Sep 23, 02:002026-09-23 02:00:00 UTC2026-09-23 03:00:00 UTC000
Sep 23, 03:002026-09-23 03:00:00 UTC2026-09-23 04:00:00 UTC101
Sep 23, 04:002026-09-23 04:00:00 UTC2026-09-23 05:00:00 UTC000
Sep 23, 05:002026-09-23 05:00:00 UTC2026-09-23 06:00:00 UTC000
Sep 23, 06:002026-09-23 06:00:00 UTC2026-09-23 07:00:00 UTC000
Sep 23, 07:002026-09-23 07:00:00 UTC2026-09-23 08:00:00 UTC000
Sep 23, 08:002026-09-23 08:00:00 UTC2026-09-23 09:00:00 UTC000
Sep 23, 09:002026-09-23 09:00:00 UTC2026-09-23 10:00:00 UTC000
Sep 23, 10:002026-09-23 10:00:00 UTC2026-09-23 11:00:00 UTC000
Sep 23, 11:002026-09-23 11:00:00 UTC2026-09-23 12:00:00 UTC101

What Hacker News says

1 sources

Show HN: JevBench, a reproducible benchmark for typed decision models

Hi HN! I built JevBench because Jev kicks ass, and the world deserves to know how the serious open source and fake lookalike projects really perform in comparison. Jev-class models return bounded choices and probabilities instead of text, and are disruptively faster and cheaper than LLMs, while being similarly intelligent on the text input they operate on. JevBench allows looking at accuracy, latency and price all at once, in a weighted way - you can even configure the weighting. A full run asks 534 English decisions. The v1.3 score combines chance-corrected Intelligence, Calibration, Speed and Cost. Leaderboard right now: #1 - Jev 74.4 #2 - SemIf 73.1 #3 - djev 73.0 #4 - Winnow-12B Q8 71.2 #5 reflex 4B 70.3. MIT harness, public items, frozen artifacts, scoring code and public per-task outc
Read the rest
omes: https://github.com/fstandhartinger/jevbench Two no-signup demos: https://who-is-right.app.mintapis.com https://is-it-ai-slop.app.mintapis.com Limitations: English-only; latency from one German server; local/demo latency gets a disclosed ×2 adjustment (+150 ms on my servers) which is an informed assumption; held-out prompts still reach evaluated services; ~1-point gaps can be noise. Wdyt?
154 points · Open discussion

MAKE YOURSELF AT HOME

A few shortcuts

/
Search the collection
?
Open this guide
Esc
Close dialogs and navigation

Your saved repositories and interests stay in this browser.