Benchmark Almanac

DAYJOB: Law

GPT-6 Astra 36.9% leads since October 2026
Not in model cards

Agents do legal assignments from a short request and a workspace of files. An agent judge applies an expert rubric, and a pass needs 95% of criteria.

Results for 15 models appear only in the Fireworks Specialized Intelligence Index. Surge had not published the tasks or task count by October 30th, 2026.

Made by
Surge AI
Released
October 2026
Kind
Law
Level
Expert
Scored
Graded by a model
Answers
Agent task
Tasks
Kept private
Strongest model tested
GPT-6 Astra #3 on the capability index
Last model added
October 2026
Scores refreshed
November 12th, 2026
16
Models scored on its own leaderboard

GPT-6 Astra leads with 36.9%

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. Every run is in the next chart.

16 models. Scroll sideways for the rest.

  1. 36.9 GPT-6 Astra
  2. 36.6 Muse Wonder
  3. 31.8 GPT-6.1 Sol
  4. 27.6 Grok 4.7
  5. 27.0 Claude Sonnet 5.5
  6. 26.5 Claude Fable 5.1
  7. 25.1 Claude Opus 5.5
  8. 23.9 Kimi K3.1 Pro
  9. 22.8 Muse Spark 1.3
  10. 16.9 Qwen 3.8 Max
  11. 9.90 Tencent Hy4 preview
  12. 9.90 Gemini 4 Argon
  13. 6.20 GLM-5.3
  14. 6.20 DeepSeek V4.1 Flash 1017
  15. 4.50 Mistral Large 4
  16. 4.20 Kimi K3
  1. 1 GPT-6 Astra max 36.9%
  2. 2 Muse Wonder max 36.6%
  3. 3 GPT-6.1 Sol max 31.8%
  4. 4 Grok 4.7 xhigh 27.6%
  5. 5 Claude Sonnet 5.5 adaptive/max 27%
  6. 6 Claude Fable 5.1 adaptive/max 26.5%
  7. 7 Claude Opus 5.5 adaptive/max 25.1%
  8. 8 Kimi K3.1 Pro 23.9%
  9. 9 Muse Spark 1.3 max 22.8%
  10. 10 Qwen 3.8 Max xhigh 16.9%
  11. 11 Tencent Hy4 preview high 9.90%
  12. 12 Gemini 4 Argon high 9.90%
  13. 13 GLM-5.3 max 6.20%
  14. 14 DeepSeek V4.1 Flash 1017 max 6.20%
  15. 15 Mistral Large 4 high 4.50%
  16. 16 Kimi K3 max 4.20%

Source: Official DAYJOB: Law leaderboard

#ModelScoreSettingDateSource
1 GPT-6 Astra 36.9% max Sep 2026 Official leaderboard
2 Muse Wonder 36.6% max Oct 2026 Official leaderboard
3 GPT-6.1 Sol 31.8% max Sep 2026 Official leaderboard
4 Grok 4.7 27.6% xhigh Sep 2026 Official leaderboard
5 Claude Sonnet 5.5 27% adaptive/max Sep 2026 Official leaderboard
6 Claude Fable 5.1 26.5% adaptive/max Sep 2026 Official leaderboard
7 Claude Opus 5.5 25.1% adaptive/max Sep 2026 Official leaderboard
8 Kimi K3.1 Pro 23.9% Oct 2026 Official leaderboard
9 Muse Spark 1.3 22.8% max Sep 2026 Official leaderboard
10 Qwen 3.8 Max 16.9% xhigh Aug 2026 Official leaderboard
11 Tencent Hy4 preview 9.90% high Aug 2026 Official leaderboard
12 Gemini 4 Argon 9.90% high Sep 2026 Official leaderboard
13 GLM-5.3 6.20% max Aug 2026 Official leaderboard
14 DeepSeek V4.1 Flash 1017 6.20% max Oct 2026 Official leaderboard
15 Mistral Large 4 4.50% high Oct 2026 Official leaderboard
16 Kimi K3 4.20% max Jul 2026 Official leaderboard

The best score, 36.9%, costs $9.26 a task

Runs on the cost frontier score higher than any run that's cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

Tencent OpenAI Meta DeepSeek Cost frontier One model at each effort level
0% 10% 20% 30% 40% $0.10 $1.00 $10 $100 Cost per task, log scale GPT-6 Astra, max GPT-6.1 Sol, max Tencent Hy4 preview, high DeepSeek V4.1 Flash 1017, max 0% 10% 20% 30% 40% $0.10 $1.00 $10 $100 Cost per task, log scale GPT-6 Astra, max GPT-6.1 Sol, max

Source: Official DAYJOB: Law leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
GPT-6 Astra max 36.9% $9.26
Muse Wonder max 36.6% $5.06
GPT-6.1 Sol max 31.8% $1.49
Grok 4.7 xhigh 27.6% $4.83
Claude Sonnet 5.5 adaptive/max 27% $14
Claude Fable 5.1 adaptive/max 26.5% $16
Claude Opus 5.5 adaptive/max 25.1% $14
Kimi K3.1 Pro 23.9% $3.92
Muse Spark 1.3 max 22.8% $1.52
Qwen 3.8 Max xhigh 16.9% $4.51
Tencent Hy4 preview high 9.90% $0.61
Gemini 4 Argon high 9.90% $3.12
GLM-5.3 max 6.20% $1.94
DeepSeek V4.1 Flash 1017 max 6.20% $0.47
Mistral Large 4 high 4.50% $2.77
Kimi K3 max 4.20% $2.52

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 Astra 36.9%

Source: Official DAYJOB: Law 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%) GPT-6 Astra Claude Fable 5.1 Qwen 3.8 Max Kimi K3 0% 20% 40% 60% 80% 100% 2027 Near the ceiling (85%) GPT-6 Astra Claude Fable 5.1

Source: Official DAYJOB: Law 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 Law, finance and medicine

All 179 in Law, finance and medicine
  • MedXpertQA-MM In use Models answer expert-level medical exam questions that combine images with clinical records and test results. Scoring is accuracy. Qwen 3.8 Max 80.4%
  • HealthBench Professional In use 525 physician-authored tasks from real clinician chats (care consult, documentation, research) with physician-adjudicated rubrics. Claude Sonnet 5.5 69.2%
  • HealthBench In use 5,000 realistic health conversations graded against physician-written rubrics. A length-adjusted variant penalizes verbosity. GPT-5 67.2%
  • FinSearchComp In use Agents search the web to fetch live financial data, look up history, and run multi-step investigations, across 635 expert-written questions. Scored by answer accuracy. Ling-3.1-flash 81.7%
  • KBL In use Models answer Korean legal knowledge and reasoning tasks and Korean bar exam questions. It has 7 knowledge tasks, 4 reasoning tasks and 2,510 bar exam items. Solar Pro 4 77.5%
  • KorMedMCQA In use Models answer 7,469 multiple-choice questions from Korean licensing exams for doctors, nurses, pharmacists and dentists, held from 2012 to 2024. Solar Pro 4 93.2%