Benchmark Almanac

MissionReport

Gemini 4 Argon 64.1% leads since October 2026
Not in model cards

Models write government status reports in Word from working files and a template, across 20 expert-written Department of Defense style assignments. Expert rubrics score each decision.

Nine models ran each task three times for 540 graded reports. Judgment and synthesis caused most lost points.

Made by
micro1
Released
October 2026
Kind
Documents and spreadsheets
Level
Expert
Scored
Graded by a model
Answers
Open-ended
Tasks
All public
Strongest model tested
Claude Fable 5.2 #1 on the capability index
Last model added
November 2026
Scores refreshed
November 12th, 2026
10
Models scored on its own leaderboard

Gemini 4 Argon leads with 64.1%

Best score per model from the benchmark's own leaderboard, read on November 12th, 2026. Where a model ran with several settings or harnesses, we keep its best.

  1. 64.1 Gemini 4 Argon
  2. 63.8 Claude Sonnet 5.5
  3. 63.8 Claude Fable 5.2
  4. 63.5 Claude Opus 5.5
  5. 63.3 Claude Fable 5.1
  6. 62.1 Muse Wonder
  7. 61.8 GPT-6 Astra
  8. 61.0 GPT-6.1 Sol
  9. 59.5 Grok 4.7
  10. 54.3 Gemini 3.8 Flash
  1. 1 Gemini 4 Argon 64.1%
  2. 2 Claude Sonnet 5.5 63.8%
  3. 3 Claude Fable 5.2 63.8%
  4. 4 Claude Opus 5.5 63.5%
  5. 5 Claude Fable 5.1 63.3%
  6. 6 Muse Wonder 62.1%
  7. 7 GPT-6 Astra 61.8%
  8. 8 GPT-6.1 Sol 61%
  9. 9 Grok 4.7 59.5%
  10. 10 Gemini 3.8 Flash 54.3%

Source: Official MissionReport leaderboard

#ModelScoreSettingDateSource
1 Gemini 4 Argon 64.1% Sep 2026 Official leaderboard
2 Claude Sonnet 5.5 63.8% Sep 2026 Official leaderboard
3 Claude Fable 5.2 63.8% Nov 2026 Official leaderboard
4 Claude Opus 5.5 63.5% Sep 2026 Official leaderboard
5 Claude Fable 5.1 63.3% Sep 2026 Official leaderboard
6 Muse Wonder 62.1% Oct 2026 Official leaderboard
7 GPT-6 Astra 61.8% Sep 2026 Official leaderboard
8 GPT-6.1 Sol 61% Sep 2026 Official leaderboard
9 Grok 4.7 59.5% Sep 2026 Official leaderboard
10 Gemini 3.8 Flash 54.3% Sep 2026 Official leaderboard

Number one over time

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

Google DeepMind
Own leaderboard 1 leader
Gemini 4 Argon 64.1%

Source: Official MissionReport leaderboard

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 Best so far
0% 20% 40% 60% 80% 100% 2027 Near the ceiling (85%) Gemini 4 Argon 0% 20% 40% 60% 80% 100% 2027 Near the ceiling (85%) Gemini 4 Argon

Source: Official MissionReport leaderboard

No lab model card reports it yet

We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark.

More in Economic work

All 97 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). Claude Fable 5 75.9%
  • 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%