Benchmark Almanac

DataAnalyticsBench

GPT-6 Astra 43.0 leads since October 2026
Not in model cards

Agents query a synthetic data warehouse to answer business questions. A judge model checks answers against a rubric and an answer key.

Plotly keeps the questions and warehouse private. It audits every perfect score and checks each failed answer by hand.

Made by
Plotly and Mila
Released
October 2026
Kind
Running a business
Level
University
Scored
Graded by a model
Answers
Agent task
Tasks
Kept private
Measures
Capability
Language
English
In and out
Text → Text
Interaction
Agentic
Environment
APIs and tools
Items from
Generated
Refresh
One fixed set
Run by
A host
Strongest model tested
GPT-6.2 Astra #1 on the capability index
Last model added
November 2026
Scores refreshed
November 20th, 2026
54
Entries on its own leaderboard, 51 models we track

5 models share the top score of 43.0 5 models share the top score of 43.0

Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.

  1. 1 GPT-6 Astra 43.0
  2. 2 GPT-6.1 Sol 43.0
  3. 3 Claude Fable 5.2 43.0
  4. 4 Gemini 4 Hydrogen 43.0
  5. 5 Gemini 4 Neon 43.0
  6. 6 Grok 4.7 42.0
  7. 7 Kimi K3.1 Pro 42.0
  8. 8 Gemini 4 Carbon 42.0
  9. 9 Tencent Hy4 42.0
  10. 10 Fugu Ultra v2 41.0
  11. 11 Claude Opus 5.5 41.0
  12. 12 GPT-6 Sol 41.0
  13. 13 Gemini 4 Argon 41.0
  14. 14 Muse Wonder 41.0
  15. 15 MiniMax-M3.1-Flash 41.0
  16. 16 GPT-6.2 Astra 41.0
  17. 17 Muse Spark 1.3 40.0
  18. 18 GPT-6.1 Terra 40.0
  19. 19 Particle-1 40.0
  20. 20 GPT-6.2 Sol 40.0
  21. 21 Tencent Hy4 preview 39.0
  22. 22 Claude Fable 5.1 39.0
  23. 23 Ember-1 39.0
  24. 24 Inkling 2 Small 39.0
  25. 25 Qwen 3.8 27B 38.0
  26. 26 GLM-5.3-Flash 38.0
  27. 27 GPT-6 Luna 38.0
  28. 28 Claude Sonnet 5.5 38.0
  29. 29 Muse Spark Open 38.0
  30. 30 Muse Ember 38.0
  31. 31 Nex-N2.6 38.0
  32. 32 Claude Haiku 5.5 37.0
  33. 33 Claude Sonnet 5 36.0
  34. 34 DeepSeek V4.1 Flash 36.0
  35. 35 Beam 1.0 36.0
  36. 36 Kimi K3 35.0
  37. 37 Muse Glimmer 35.0
  38. 38 Mercury 3 35.0
  39. 39 Ripple-65B Preview 35.0
  40. 40 Qwen4-Max 35.0
  41. 41 Nemotron 3 Ultra 34.0
  42. 42 Mistral Large 4 32.0
  43. 43 Inkling-Small 30.0
  44. 44 Nemotron 3.5 Super VL 30.0
  45. 45 Nemotron 3 Super 29.0
  46. 46 Inkling 29.0
  47. 47 Solar Pro 4 28.0
  48. 48 gpt-oss-120b 27.0
  49. 49 Claude Haiku 4.5 27.0
  50. 50 Gemma 4 31B IT 26.0
  51. 51 Mistral Medium 3.5 25.0
  52. 52 Solar Mini 4 21.0
  53. 53 Nemotron 3.5 Lightning 30B-A3B 19.0
  54. 54 Command A+ 18.0

Source: Official DataAnalyticsBench leaderboard

Source: Official DataAnalyticsBench leaderboard

#EntryScoreSettingReleasedRun onSource
1 GPT-6 Astra 43.0 Sep 2026 Sep 2026 Official leaderboard
2 GPT-6.1 Sol 43.0 Sep 2026 Sep 2026 Official leaderboard
3 Claude Fable 5.2 43.0 Nov 2026 Nov 2026 Official leaderboard
4 Gemini 4 Hydrogen 43.0 Nov 2026 Nov 2026 Official leaderboard
5 Gemini 4 Neon 43.0 Nov 2026 Nov 2026 Official leaderboard
6 Grok 4.7 42.0 Sep 2026 Sep 2026 Official leaderboard
7 Kimi K3.1 Pro 42.0 Oct 2026 Oct 2026 Official leaderboard
8 Gemini 4 Carbon 42.0 Oct 2026 Nov 2026 Official leaderboard
9 Tencent Hy4 42.0 Nov 2026 Nov 2026 Official leaderboard
10 Fugu Ultra v2 41.0 Sep 2026 Official leaderboard
11 Claude Opus 5.5 41.0 Sep 2026 Sep 2026 Official leaderboard
12 GPT-6 Sol 41.0 Sep 2026 Sep 2026 Official leaderboard
13 Gemini 4 Argon 41.0 Sep 2026 Oct 2026 Official leaderboard
14 Muse Wonder 41.0 Oct 2026 Oct 2026 Official leaderboard
15 MiniMax-M3.1-Flash 41.0 Oct 2026 Oct 2026 Official leaderboard
16 GPT-6.2 Astra 41.0 Nov 2026 Nov 2026 Official leaderboard
17 Muse Spark 1.3 40.0 Sep 2026 Sep 2026 Official leaderboard
18 GPT-6.1 Terra 40.0 Oct 2026 Oct 2026 Official leaderboard
19 Particle-1 40.0 Oct 2026 Official leaderboard
20 GPT-6.2 Sol 40.0 Nov 2026 Nov 2026 Official leaderboard
21 Tencent Hy4 preview 39.0 Aug 2026 Nov 2026 Official leaderboard
22 Claude Fable 5.1 39.0 Sep 2026 Sep 2026 Official leaderboard
23 Ember-1 39.0 Sep 2026 Sep 2026 Official leaderboard
24 Inkling 2 Small 39.0 Nov 2026 Nov 2026 Official leaderboard
25 Qwen 3.8 27B 38.0 Aug 2026 Aug 2026 Official leaderboard
26 GLM-5.3-Flash 38.0 Aug 2026 Aug 2026 Official leaderboard
27 GPT-6 Luna 38.0 Sep 2026 Sep 2026 Official leaderboard
28 Claude Sonnet 5.5 38.0 Sep 2026 Sep 2026 Official leaderboard
29 Muse Spark Open 38.0 Oct 2026 Oct 2026 Official leaderboard
30 Muse Ember 38.0 Nov 2026 Nov 2026 Official leaderboard
31 Nex-N2.6 38.0 Nov 2026 Official leaderboard
32 Claude Haiku 5.5 37.0 Oct 2026 Oct 2026 Official leaderboard
33 Claude Sonnet 5 36.0 Jun 2026 Jun 2026 Official leaderboard
34 DeepSeek V4.1 Flash 36.0 Sep 2026 Sep 2026 Official leaderboard
35 Beam 1.0 36.0 Oct 2026 Oct 2026 Official leaderboard
36 Kimi K3 35.0 Jul 2026 Jul 2026 Official leaderboard
37 Muse Glimmer 35.0 Aug 2026 Aug 2026 Official leaderboard
38 Mercury 3 35.0 Oct 2026 Oct 2026 Official leaderboard
39 Ripple-65B Preview 35.0 Nov 2026 Nov 2026 Official leaderboard
40 Qwen4-Max 35.0 Nov 2026 Nov 2026 Official leaderboard
41 Nemotron 3 Ultra 34.0 Jun 2026 Jun 2026 Official leaderboard
42 Mistral Large 4 32.0 Oct 2026 Oct 2026 Official leaderboard
43 Inkling-Small 30.0 Jul 2026 Jul 2026 Official leaderboard
44 Nemotron 3.5 Super VL 30.0 Oct 2026 Oct 2026 Official leaderboard
45 Nemotron 3 Super 29.0 Mar 2026 Mar 2026 Official leaderboard
46 Inkling 29.0 Jul 2026 Jul 2026 Official leaderboard
47 Solar Pro 4 28.0 Aug 2026 Aug 2026 Official leaderboard
48 gpt-oss-120b 27.0 Aug 2025 Aug 2025 Official leaderboard
49 Claude Haiku 4.5 27.0 Oct 2025 Oct 2025 Official leaderboard
50 Gemma 4 31B IT 26.0 Apr 2026 Apr 2026 Official leaderboard
51 Mistral Medium 3.5 25.0 Apr 2026 Apr 2026 Official leaderboard
52 Solar Mini 4 21.0 Oct 2026 Sep 2026 Official leaderboard
53 Nemotron 3.5 Lightning 30B-A3B 19.0 Aug 2026 Aug 2026 Official leaderboard
54 Command A+ 18.0 May 2026 Sep 2026 Official leaderboard
#ModelScoreSettingReleasedRun onSource
1 GPT-6 Astra 43.0 Sep 2026 Sep 2026 Official leaderboard
2 GPT-6.1 Sol 43.0 Sep 2026 Sep 2026 Official leaderboard
3 Claude Fable 5.2 43.0 Nov 2026 Nov 2026 Official leaderboard
4 Gemini 4 Hydrogen 43.0 Nov 2026 Nov 2026 Official leaderboard
5 Gemini 4 Neon 43.0 Nov 2026 Nov 2026 Official leaderboard
6 Grok 4.7 42.0 Sep 2026 Sep 2026 Official leaderboard
7 Kimi K3.1 Pro 42.0 Oct 2026 Oct 2026 Official leaderboard
8 Gemini 4 Carbon 42.0 Oct 2026 Nov 2026 Official leaderboard
9 Tencent Hy4 42.0 Nov 2026 Nov 2026 Official leaderboard
10 Claude Opus 5.5 41.0 Sep 2026 Sep 2026 Official leaderboard
11 GPT-6 Sol 41.0 Sep 2026 Sep 2026 Official leaderboard
12 Gemini 4 Argon 41.0 Sep 2026 Oct 2026 Official leaderboard
13 Muse Wonder 41.0 Oct 2026 Oct 2026 Official leaderboard
14 MiniMax-M3.1-Flash 41.0 Oct 2026 Oct 2026 Official leaderboard
15 GPT-6.2 Astra 41.0 Nov 2026 Nov 2026 Official leaderboard
16 Muse Spark 1.3 40.0 Sep 2026 Sep 2026 Official leaderboard
17 GPT-6.1 Terra 40.0 Oct 2026 Oct 2026 Official leaderboard
18 GPT-6.2 Sol 40.0 Nov 2026 Nov 2026 Official leaderboard
19 Tencent Hy4 preview 39.0 Aug 2026 Nov 2026 Official leaderboard
20 Claude Fable 5.1 39.0 Sep 2026 Sep 2026 Official leaderboard
21 Ember-1 39.0 Sep 2026 Sep 2026 Official leaderboard
22 Inkling 2 Small 39.0 Nov 2026 Nov 2026 Official leaderboard
23 Qwen 3.8 27B 38.0 Aug 2026 Aug 2026 Official leaderboard
24 GLM-5.3-Flash 38.0 Aug 2026 Aug 2026 Official leaderboard
25 GPT-6 Luna 38.0 Sep 2026 Sep 2026 Official leaderboard
26 Claude Sonnet 5.5 38.0 Sep 2026 Sep 2026 Official leaderboard
27 Muse Spark Open 38.0 Oct 2026 Oct 2026 Official leaderboard
28 Muse Ember 38.0 Nov 2026 Nov 2026 Official leaderboard
29 Claude Haiku 5.5 37.0 Oct 2026 Oct 2026 Official leaderboard
30 Claude Sonnet 5 36.0 Jun 2026 Jun 2026 Official leaderboard
31 DeepSeek V4.1 Flash 36.0 Sep 2026 Sep 2026 Official leaderboard
32 Beam 1.0 36.0 Oct 2026 Oct 2026 Official leaderboard
33 Kimi K3 35.0 Jul 2026 Jul 2026 Official leaderboard
34 Muse Glimmer 35.0 Aug 2026 Aug 2026 Official leaderboard
35 Mercury 3 35.0 Oct 2026 Oct 2026 Official leaderboard
36 Ripple-65B Preview 35.0 Nov 2026 Nov 2026 Official leaderboard
37 Qwen4-Max 35.0 Nov 2026 Nov 2026 Official leaderboard
38 Nemotron 3 Ultra 34.0 Jun 2026 Jun 2026 Official leaderboard
39 Mistral Large 4 32.0 Oct 2026 Oct 2026 Official leaderboard
40 Inkling-Small 30.0 Jul 2026 Jul 2026 Official leaderboard
41 Nemotron 3.5 Super VL 30.0 Oct 2026 Oct 2026 Official leaderboard
42 Nemotron 3 Super 29.0 Mar 2026 Mar 2026 Official leaderboard
43 Inkling 29.0 Jul 2026 Jul 2026 Official leaderboard
44 Solar Pro 4 28.0 Aug 2026 Aug 2026 Official leaderboard
45 gpt-oss-120b 27.0 Aug 2025 Aug 2025 Official leaderboard
46 Claude Haiku 4.5 27.0 Oct 2025 Oct 2025 Official leaderboard
47 Gemma 4 31B IT 26.0 Apr 2026 Apr 2026 Official leaderboard
48 Mistral Medium 3.5 25.0 Apr 2026 Apr 2026 Official leaderboard
49 Solar Mini 4 21.0 Oct 2026 Sep 2026 Official leaderboard
50 Nemotron 3.5 Lightning 30B-A3B 19.0 Aug 2026 Aug 2026 Official leaderboard
51 Command A+ 18.0 May 2026 Sep 2026 Official leaderboard

Score against run cost

The article compares the 43-question score with the median cost of one complete run. The line joins the runs that no other run beats on both measures. We redrew this chart from the numbers on the official leaderboard.

20 30 40 50 $0.1$1$10$100 Cost per run ($), log scale, lower is better Questions correct (score) GPT-6.1 SolTencent Hy4Particle-1Tencent Hy4 preview 20 30 40 50 $0.1$1$10$100 Cost per run ($), log scale, lower is better Questions correct (score) GPT-6.1 SolTencent Hy4Particle-1Tencent Hy4 preview
ModelSettingQuestions correctCost per run
GPT-6.1 Sol 43 score $2.61
Gemini 4 Hydrogen 43 score $6.36
Gemini 4 Neon 43 score $17.03
GPT-6 Astra 43 score $19.97
Claude Fable 5.2 43 score $35.74
Tencent Hy4 42 score $2.11
Grok 4.7 42 score $7.85
Gemini 4 Carbon 42 score $11.55
Kimi K3.1 Pro 42 score $13.47
GPT-6 Sol 41 score $2.61
MiniMax-M3.1-Flash 41 score $4.09
GPT-6.2 Astra 41 score $4.11
Claude Opus 5.5 41 score $13.31
Gemini 4 Argon 41 score $13.47
Muse Wonder 41 score $16.98
Fugu Ultra v2 41 score $49.06
Particle-1 40 score $0.35
GPT-6.2 Sol 40 score $1.22
GPT-6.1 Terra 40 score $1.31
Muse Spark 1.3 40 score $9.48
Tencent Hy4 preview 39 score $0.13
Inkling 2 Small 39 score $1.81
Ember-1 39 score $5.66
Claude Fable 5.1 39 score $31.49
GPT-6 Luna 38 score $0.13
GLM-5.3-Flash 38 score $0.69
Muse Ember 38 score $1.04
Qwen 3.8 27B 38 score $1.71
Muse Spark Open 38 score $5.16
Claude Sonnet 5.5 38 score $7.3
Nex-N2.6 38 score $10.19
Claude Haiku 5.5 37 score $0.38
DeepSeek V4.1 Flash 36 score $0.46
Beam 1.0 36 score $8.29
Claude Sonnet 5 36 score $8.67
Mercury 3 35 score $0.41
Ripple-65B Preview 35 score $2.21
Qwen4-Max 35 score $2.29
Muse Glimmer 35 score $3.18
Kimi K3 35 score $5.01
Nemotron 3 Ultra 34 score $7.06
Mistral Large 4 32 score $1.94
Nemotron 3.5 Super VL 30 score $0.46
Inkling-Small 30 score $1.93
Nemotron 3 Super 29 score $0.77
Inkling 29 score $9.54
Solar Pro 4 28 score $2.04
gpt-oss-120b 27 score $0.22
Claude Haiku 4.5 27 score $3.27
Gemma 4 31B IT 26 score $0.62
Mistral Medium 3.5 25 score $20.51
Solar Mini 4 21 score $1.04
Nemotron 3.5 Lightning 30B-A3B 19 score $0.94
Command A+ 18 score $19.47

Score against run cost

The article compares the 43-question score with the median cost of one complete run. The line joins the runs that no other run beats on both measures. We redrew this chart from the numbers on the official leaderboard.

20 30 40 50 $0.1$1$10$100 Cost per run ($), log scale, lower is better Questions correct (score) GPT-6.1 SolTencent Hy4GPT-6.2 SolTencent Hy4 preview 20 30 40 50 $0.1$1$10$100 Cost per run ($), log scale, lower is better Questions correct (score) GPT-6.1 SolTencent Hy4GPT-6.2 SolTencent Hy4 preview
ModelSettingQuestions correctCost per run
GPT-6.1 Sol 43 score $2.61
Gemini 4 Hydrogen 43 score $6.36
Gemini 4 Neon 43 score $17.03
GPT-6 Astra 43 score $19.97
Claude Fable 5.2 43 score $35.74
Tencent Hy4 42 score $2.11
Grok 4.7 42 score $7.85
Gemini 4 Carbon 42 score $11.55
Kimi K3.1 Pro 42 score $13.47
GPT-6 Sol 41 score $2.61
MiniMax-M3.1-Flash 41 score $4.09
GPT-6.2 Astra 41 score $4.11
Claude Opus 5.5 41 score $13.31
Gemini 4 Argon 41 score $13.47
Muse Wonder 41 score $16.98
GPT-6.2 Sol 40 score $1.22
GPT-6.1 Terra 40 score $1.31
Muse Spark 1.3 40 score $9.48
Tencent Hy4 preview 39 score $0.13
Inkling 2 Small 39 score $1.81
Ember-1 39 score $5.66
Claude Fable 5.1 39 score $31.49
GPT-6 Luna 38 score $0.13
GLM-5.3-Flash 38 score $0.69
Muse Ember 38 score $1.04
Qwen 3.8 27B 38 score $1.71
Muse Spark Open 38 score $5.16
Claude Sonnet 5.5 38 score $7.3
Claude Haiku 5.5 37 score $0.38
DeepSeek V4.1 Flash 36 score $0.46
Beam 1.0 36 score $8.29
Claude Sonnet 5 36 score $8.67
Mercury 3 35 score $0.41
Ripple-65B Preview 35 score $2.21
Qwen4-Max 35 score $2.29
Muse Glimmer 35 score $3.18
Kimi K3 35 score $5.01
Nemotron 3 Ultra 34 score $7.06
Mistral Large 4 32 score $1.94
Nemotron 3.5 Super VL 30 score $0.46
Inkling-Small 30 score $1.93
Nemotron 3 Super 29 score $0.77
Inkling 29 score $9.54
Solar Pro 4 28 score $2.04
gpt-oss-120b 27 score $0.22
Claude Haiku 4.5 27 score $3.27
Gemma 4 31B IT 26 score $0.62
Mistral Medium 3.5 25 score $20.51
Solar Mini 4 21 score $1.04
Nemotron 3.5 Lightning 30B-A3B 19 score $0.94
Command A+ 18 score $19.47

The best score, 43.0, costs $2.61 for the full run

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 Cost frontier One model at each effort level
0.00 10.0 20.0 30.0 40.0 50.0 $0.10 $1.00 $10 $100 Cost of the full run, log scale GPT-6.1 Sol GPT-6.2 Sol Tencent Hy4 preview 0.00 10.0 20.0 30.0 40.0 50.0 $0.10 $1.00 $10 $100 Cost of the full run, log scale GPT-6.1 Sol GPT-6.2 Sol

Source: Official DataAnalyticsBench leaderboard

ModelEffortHarnessScoreCost of the full runOutput tokens
GPT-6 Astra 43.0 $20
GPT-6.1 Sol 43.0 $2.61
Claude Fable 5.2 43.0 $36
Gemini 4 Hydrogen 43.0 $6.36
Gemini 4 Neon 43.0 $17
Grok 4.7 42.0 $7.85
Kimi K3.1 Pro 42.0 $13
Gemini 4 Carbon 42.0 $12
Tencent Hy4 42.0 $2.11
Claude Opus 5.5 41.0 $13
GPT-6 Sol 41.0 $2.61
Gemini 4 Argon 41.0 $13
Muse Wonder 41.0 $17
MiniMax-M3.1-Flash 41.0 $4.09
GPT-6.2 Astra 41.0 $4.11
Muse Spark 1.3 40.0 $9.48
GPT-6.1 Terra 40.0 $1.31
GPT-6.2 Sol 40.0 $1.22
Tencent Hy4 preview 39.0 $0.13
Claude Fable 5.1 39.0 $31
Ember-1 39.0 $5.66
Inkling 2 Small 39.0 $1.81
Qwen 3.8 27B 38.0 $1.71
GLM-5.3-Flash 38.0 $0.69
GPT-6 Luna 38.0 $0.13
Claude Sonnet 5.5 38.0 $7.30
Muse Spark Open 38.0 $5.16
Muse Ember 38.0 $1.04
Claude Haiku 5.5 37.0 $0.38
Claude Sonnet 5 36.0 $8.67
DeepSeek V4.1 Flash 36.0 $0.46
Beam 1.0 36.0 $8.29
Kimi K3 35.0 $5.01
Muse Glimmer 35.0 $3.18
Mercury 3 35.0 $0.41
Ripple-65B Preview 35.0 $2.21
Qwen4-Max 35.0 $2.29
Nemotron 3 Ultra 34.0 $7.06
Mistral Large 4 32.0 $1.94
Inkling-Small 30.0 $1.93
Nemotron 3.5 Super VL 30.0 $0.46
Nemotron 3 Super 29.0 $0.77
Inkling 29.0 $9.54
Solar Pro 4 28.0 $2.04
gpt-oss-120b 27.0 $0.22
Claude Haiku 4.5 27.0 $3.27
Gemma 4 31B IT 26.0 $0.62
Mistral Medium 3.5 25.0 $21
Solar Mini 4 21.0 $1.04
Nemotron 3.5 Lightning 30B-A3B 19.0 $0.94
Command A+ 18.0 $19

The best score, 43.0, costs $2.61 for the full run

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 Cost frontier One model at each effort level
0.00 10.0 20.0 30.0 40.0 50.0 $0.10 $1.00 $10 $100 Cost of the full run, log scale GPT-6.1 Sol GPT-6.2 Sol Tencent Hy4 preview 0.00 10.0 20.0 30.0 40.0 50.0 $0.10 $1.00 $10 $100 Cost of the full run, log scale GPT-6.1 Sol GPT-6.2 Sol

Source: Official DataAnalyticsBench leaderboard

ModelEffortHarnessScoreCost of the full runOutput tokens
GPT-6 Astra 43.0 $20
GPT-6.1 Sol 43.0 $2.61
Claude Fable 5.2 43.0 $36
Gemini 4 Hydrogen 43.0 $6.36
Gemini 4 Neon 43.0 $17
Grok 4.7 42.0 $7.85
Kimi K3.1 Pro 42.0 $13
Gemini 4 Carbon 42.0 $12
Tencent Hy4 42.0 $2.11
Claude Opus 5.5 41.0 $13
GPT-6 Sol 41.0 $2.61
Gemini 4 Argon 41.0 $13
Muse Wonder 41.0 $17
MiniMax-M3.1-Flash 41.0 $4.09
GPT-6.2 Astra 41.0 $4.11
Muse Spark 1.3 40.0 $9.48
GPT-6.1 Terra 40.0 $1.31
GPT-6.2 Sol 40.0 $1.22
Tencent Hy4 preview 39.0 $0.13
Claude Fable 5.1 39.0 $31
Ember-1 39.0 $5.66
Inkling 2 Small 39.0 $1.81
Qwen 3.8 27B 38.0 $1.71
GLM-5.3-Flash 38.0 $0.69
GPT-6 Luna 38.0 $0.13
Claude Sonnet 5.5 38.0 $7.30
Muse Spark Open 38.0 $5.16
Muse Ember 38.0 $1.04
Claude Haiku 5.5 37.0 $0.38
Claude Sonnet 5 36.0 $8.67
DeepSeek V4.1 Flash 36.0 $0.46
Beam 1.0 36.0 $8.29
Kimi K3 35.0 $5.01
Muse Glimmer 35.0 $3.18
Mercury 3 35.0 $0.41
Ripple-65B Preview 35.0 $2.21
Qwen4-Max 35.0 $2.29
Nemotron 3 Ultra 34.0 $7.06
Mistral Large 4 32.0 $1.94
Inkling-Small 30.0 $1.93
Nemotron 3.5 Super VL 30.0 $0.46
Nemotron 3 Super 29.0 $0.77
Inkling 29.0 $9.54
Solar Pro 4 28.0 $2.04
gpt-oss-120b 27.0 $0.22
Claude Haiku 4.5 27.0 $3.27
Gemma 4 31B IT 26.0 $0.62
Mistral Medium 3.5 25.0 $21
Solar Mini 4 21.0 $1.04
Nemotron 3.5 Lightning 30B-A3B 19.0 $0.94
Command A+ 18.0 $19

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

Source: Official DataAnalyticsBench leaderboard

Scores over time

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

Plot with other benchmarks
On the benchmark’s own leaderboard
20.0 30.0 40.0 2026 2027 20.0 30.0 40.0 2026 2027

Source: Official DataAnalyticsBench leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Economic work

All 210 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
  • Agents' Last Exam v1.1 In use Agents complete 122 public long-horizon professional workflow tasks across 55 sub-industries. Deterministic graders check outcomes against hidden references. GPT-6.2 Astra 47.5%
  • 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%

Data audit

All 10 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access The tasks are private.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (54 entries).
  • Pass Models scored 51 models we track, plus 3 other entries on its leaderboard.
  • Pass Newest score The newest is from November 2026, for Gemini 4 Neon.
  • Pass Frontier models 10 of the top 10 have a score, led by GPT-6.2 Astra (#1).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on plotly.com: 51 rows for models we track and 3 other entries at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check