Benchmark Almanac

Short Form Bench

GPT-6.1 Sol $2,949 leads since Oct 2026 Not in model cards

Agents act as growth lead of a small brand for a simulated week. They post short videos, run the store and answer customers, sponsors and the founder.

Some simulated parties make mistakes in the brand's favor to test honesty. Scored by business value in US dollars. 9 models in the Fireworks index.

Made by Naïve Released Oct 2026 Simulated companies Tasks kept private Last model added Sep 2026 Scores refreshed 2 Nov 2026
Level
Expert Specialists with years of training, often over hours of work
Scored
Settled by events Predictions are scored against what later happens.
Answers
Agent task Many steps in a terminal, browser, desktop, app, game or robot.
Leaderboard usenaive.ai
10
Entries on its own leaderboard, 9 models we track

GPT-6.1 Sol leads with $2,949

Best score per entry from the benchmark's own leaderboard, read on 2 Nov 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.

  1. $2,949 GPT-6.1 Sol
  2. $2,803 GPT-6 Astra
  3. $2,578 Grok 4.7
  4. $2,410 GPT-5.6 Sol
  5. $2,394 GPT-6 Sol
  6. $2,389 Claude Opus 5.5
  7. $2,355 Claude Sonnet 5.5
  8. $2,107 GPT-6 Luna
  9. $1,940 Gemini 3.8 Flash
  10. $1,793 Do nothing (baseline)
  1. 1 GPT-6.1 Sol $2,949
  2. 2 GPT-6 Astra $2,803
  3. 3 Grok 4.7 $2,578
  4. 4 GPT-5.6 Sol $2,410
  5. 5 GPT-6 Sol $2,394
  6. 6 Claude Opus 5.5 $2,389
  7. 7 Claude Sonnet 5.5 $2,355
  8. 8 GPT-6 Luna $2,107
  9. 9 Gemini 3.8 Flash $1,940
  10. 10 Do nothing (baseline) $1,793
#EntryScoreSettingDateSource
1 GPT-6.1 Sol $2,949 Sep 2026 Official leaderboard
2 GPT-6 Astra $2,803 Sep 2026 Official leaderboard
3 Grok 4.7 $2,578 Sep 2026 Official leaderboard
4 GPT-5.6 Sol $2,410 Jul 2026 Official leaderboard
5 GPT-6 Sol $2,394 Sep 2026 Official leaderboard
6 Claude Opus 5.5 $2,389 Sep 2026 Official leaderboard
7 Claude Sonnet 5.5 $2,355 Sep 2026 Official leaderboard
8 GPT-6 Luna $2,107 Sep 2026 Official leaderboard
9 Gemini 3.8 Flash $1,940 Sep 2026 Official leaderboard
10 Do nothing (baseline) $1,793 Official leaderboard

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

OpenAI
Own leaderboard 1 leader
GPT-6.1 Sol $2,949

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

On the benchmark’s own leaderboard
$1,750 $2,000 $2,250 $2,500 $2,750 $3,000 $1,750 $2,000 $2,250 $2,500 $2,750 $3,000

No lab model card reports it yet

We read the headline results of 479 launch posts from 51 labs, and none of them list this benchmark.

More in Economic work

All 91 in Economic work
  • JobBench In use 130 tasks from 35 occupations workers most want delegated. Agent works through messy reference files, graded by fact-anchored binary rubrics. GLM-5.3 61.4%
  • GDPval In use Real work tasks from 44 occupations with reference files. Experts blindly compare model deliverables with expert work. Metric is win-or-tie rate. Seed2.1 Pro 87.9%
  • GDPval-AA v2.1 In use Agents with shell and web access complete 220 GDPval gold tasks from 44 occupations. Blind pairwise comparisons give Elo ratings. Claude Opus 5.5 1,846
  • AA-Briefcase v1.1 In use Version 1.1 of Artificial Analysis's multi-hour office work evaluation, with Elo ratings fitted by Crowd-BT. Claude Opus 5.5 1,822
  • CoWorkBench In use Agent completes long-horizon productivity tasks across computer science, finance, law and medicine. Qwen in-house scoring (method unpublished). Qwen3.8-Omni-Flash 75.3%
  • OfficeQA In use Grounded reasoning over a large corpus of U.S. Treasury Bulletin documents: locate tables and compute precise numeric answers. Claude Opus 4.7 86.3%