Benchmark Almanac

Vals Finance Index

Gemini 4 Carbon 67% leads since October 2026
Gemini 4 Carbon 67% leads since October 2026
In use

Agents are scored on Finance Agent v2.1, Excel Modeling Benchmark v1.1 and Tax Agent Bench, combined with weights from each sub-sector share of US GDP.

Securities work (1.6% of GDP) averages Finance Agent and EMB. Tax work (0.5%) uses Tax Agent Bench all-pass accuracy. 35 models on October 29th, 2026.

Made by
Vals AI
Released
October 2026
Kind
Composite work indices
Level
Expert
Scored
Graded by a model
Answers
Agent task
Tasks
Kept private
Strongest model tested
Claude Fable 5.2 #1 on the capability index
Last in a launch post
November 2026
Last model added
November 2026
Scores refreshed
November 13th, 2026
40
Models scored on its own leaderboard
1
Launch post lists it, from 1 lab

Gemini 4 Carbon leads with 67%

Best score per model from the benchmark's own leaderboard, read on November 13th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart.

Gemini 4 Carbon leads with 67%, claims included

Best score per model from every source, including labs’ own numbers and claims Epoch AI collected from reports and papers. Labs pick their own settings, so a claim can beat an independent run of the same model.

40 models. Scroll sideways for the rest.

  1. 67.0 Gemini 4 Carbon
  2. 66.1 Gemini 4 Hydrogen
  3. 66.0 Gemini 4 Argon
  4. 65.3 Claude Opus 5.5
  5. 64.5 Muse Wonder Full
  6. 64.5 Claude Sonnet 5.5
  7. 63.6 Claude Fable 5.2
  8. 61.7 Qwen3.8 Max (0902)
  9. 61.5 Inkling 2 Small
  10. 61.4 GLM-5.3
  11. 61.2 Muse Wonder
  12. 60.9 Muse Ember
  13. 60.9 DeepSeek V4.1 Flash 1017
  14. 60.5 Gemini 3.8 Flash
  15. 58.8 Grok 4.7
  16. 58.0 Tencent Hy4
  17. 58.0 Kimi K3
  18. 57.9 MiMo-V2.6 Pro
  19. 57.5 Tencent Hy4 preview
  20. 57.5 Claude Haiku 5.5
  21. 57.1 MiMo-V2.6 Flash
  22. 57.1 DeepSeek V4.1 Flash
  23. 57.0 Mistral Large 4
  24. 56.9 Mistral Large 4 Preview
  25. 56.0 Kimi K3.1 Pro
  26. 55.9 MiniMax-M3.1-Flash
  27. 55.8 GLM-5.3-Flash
  28. 55.8 GPT-6.1 Sol
  29. 55.6 GPT-6 Astra
  30. 55.4 Step 5 Preview
  31. 53.3 GPT-6.1 Terra
  32. 53.0 GPT-6 Luna
  33. 53.0 Muse Spark Open
  34. 52.9 DeepSeek V4 Pro 0813
  35. 48.4 MiniMax-M3
  36. 47.2 Mercury 3
  37. 46.5 Beam 1.0
  38. 41.0 Nemotron 3.5 Super VL
  39. 38.3 Inkling
  40. 36.3 Inkling-Small
  1. 1 Gemini 4 Carbon 67%
  2. 2 Gemini 4 Hydrogen 66.1%
  3. 3 Gemini 4 Argon 66%
  4. 4 Claude Opus 5.5 65.3%
  5. 5 Muse Wonder Full 64.5%
  6. 6 Claude Sonnet 5.5 64.5%
  7. 7 Claude Fable 5.2 63.6%
  8. 8 Qwen3.8 Max (0902) 61.7%
  9. 9 Inkling 2 Small 61.5%
  10. 10 GLM-5.3 61.4%
  11. 11 Muse Wonder 61.2%
  12. 12 Muse Ember 60.9%
  13. 13 DeepSeek V4.1 Flash 1017 60.9%
  14. 14 Gemini 3.8 Flash 60.5%
  15. 15 Grok 4.7 58.8%
  16. 16 Tencent Hy4 58%
  17. 17 Kimi K3 58%
  18. 18 MiMo-V2.6 Pro 57.9%
  19. 19 Tencent Hy4 preview 57.5%
  20. 20 Claude Haiku 5.5 57.5%
  21. 21 MiMo-V2.6 Flash 57.1%
  22. 22 DeepSeek V4.1 Flash 57.1%
  23. 23 Mistral Large 4 57%
  24. 24 Mistral Large 4 Preview 56.9%
  25. 25 Kimi K3.1 Pro 56%
  26. 26 MiniMax-M3.1-Flash 55.9%
  27. 27 GLM-5.3-Flash 55.8%
  28. 28 GPT-6.1 Sol 55.8%
  29. 29 GPT-6 Astra 55.6%
  30. 30 Step 5 Preview 55.4%
  31. 31 GPT-6.1 Terra 53.3%
  32. 32 GPT-6 Luna 53%
  33. 33 Muse Spark Open 53%
  34. 34 DeepSeek V4 Pro 0813 52.9%
  35. 35 MiniMax-M3 48.4%
  36. 36 Mercury 3 47.2%
  37. 37 Beam 1.0 46.5%
  38. 38 Nemotron 3.5 Super VL 41%
  39. 39 Inkling 38.3%
  40. 40 Inkling-Small 36.3%

Source: Official Vals Finance Index leaderboard

40 models. Scroll sideways for the rest.

  1. 67.0 Gemini 4 Carbon
  2. 66.1 Gemini 4 Hydrogen
  3. 66.0 Gemini 4 Argon
  4. 65.3 Claude Opus 5.5
  5. 64.5 Muse Wonder Full
  6. 64.5 Claude Sonnet 5.5
  7. 63.6 Claude Fable 5.2
  8. 61.7 Qwen3.8 Max (0902)
  9. 61.5 Inkling 2 Small
  10. 61.4 GLM-5.3
  11. 61.2 Muse Wonder
  12. 60.9 Muse Ember
  13. 60.9 DeepSeek V4.1 Flash 1017
  14. 60.5 Gemini 3.8 Flash
  15. 58.8 Grok 4.7
  16. 58.0 Tencent Hy4
  17. 58.0 Kimi K3
  18. 57.9 MiMo-V2.6 Pro
  19. 57.5 Tencent Hy4 preview
  20. 57.5 Claude Haiku 5.5
  21. 57.1 MiMo-V2.6 Flash
  22. 57.1 DeepSeek V4.1 Flash
  23. 57.0 Mistral Large 4
  24. 56.9 Mistral Large 4 Preview
  25. 56.0 Kimi K3.1 Pro
  26. 55.9 MiniMax-M3.1-Flash
  27. 55.8 GLM-5.3-Flash
  28. 55.8 GPT-6.1 Sol
  29. 55.6 GPT-6 Astra
  30. 55.4 Step 5 Preview
  31. 53.3 GPT-6.1 Terra
  32. 53.0 GPT-6 Luna
  33. 53.0 Muse Spark Open
  34. 52.9 DeepSeek V4 Pro 0813
  35. 48.4 MiniMax-M3
  36. 47.2 Mercury 3
  37. 46.5 Beam 1.0
  38. 41.0 Nemotron 3.5 Super VL
  39. 38.3 Inkling
  40. 36.3 Inkling-Small
  1. 1 Gemini 4 Carbon 67%
  2. 2 Gemini 4 Hydrogen 66.1%
  3. 3 Gemini 4 Argon 66%
  4. 4 Claude Opus 5.5 65.3%
  5. 5 Muse Wonder Full 64.5%
  6. 6 Claude Sonnet 5.5 64.5%
  7. 7 Claude Fable 5.2 63.6%
  8. 8 Qwen3.8 Max (0902) 61.7%
  9. 9 Inkling 2 Small 61.5%
  10. 10 GLM-5.3 61.4%
  11. 11 Muse Wonder 61.2%
  12. 12 Muse Ember 60.9%
  13. 13 DeepSeek V4.1 Flash 1017 60.9%
  14. 14 Gemini 3.8 Flash 60.5%
  15. 15 Grok 4.7 58.8%
  16. 16 Tencent Hy4 58%
  17. 17 Kimi K3 58%
  18. 18 MiMo-V2.6 Pro 57.9%
  19. 19 Tencent Hy4 preview 57.5%
  20. 20 Claude Haiku 5.5 57.5%
  21. 21 MiMo-V2.6 Flash 57.1%
  22. 22 DeepSeek V4.1 Flash 57.1%
  23. 23 Mistral Large 4 57%
  24. 24 Mistral Large 4 Preview 56.9%
  25. 25 Kimi K3.1 Pro 56%
  26. 26 MiniMax-M3.1-Flash 55.9%
  27. 27 GLM-5.3-Flash 55.8%
  28. 28 GPT-6.1 Sol 55.8%
  29. 29 GPT-6 Astra 55.6%
  30. 30 Step 5 Preview 55.4%
  31. 31 GPT-6.1 Terra 53.3%
  32. 32 GPT-6 Luna 53%
  33. 33 Muse Spark Open 53%
  34. 34 DeepSeek V4 Pro 0813 52.9%
  35. 35 MiniMax-M3 48.4%
  36. 36 Mercury 3 47.2%
  37. 37 Beam 1.0 46.5%
  38. 38 Nemotron 3.5 Super VL 41%
  39. 39 Inkling 38.3%
  40. 40 Inkling-Small 36.3%

Sources: Official Vals Finance Index leaderboard and labs’ model cards

#ModelScoreSettingDateSource
1 Gemini 4 Carbon 67% Oct 2026 Official leaderboard
2 Gemini 4 Hydrogen 66.1% Nov 2026 Official leaderboard
3 Gemini 4 Argon 66% Sep 2026 Official leaderboard
4 Claude Opus 5.5 65.3% Sep 2026 Official leaderboard
5 Muse Wonder Full 64.5% Oct 2026 Official leaderboard
6 Claude Sonnet 5.5 64.5% Sep 2026 Official leaderboard
7 Claude Fable 5.2 63.6% Nov 2026 Official leaderboard
8 Qwen3.8 Max (0902) 61.7% Sep 2026 Official leaderboard
9 Inkling 2 Small 61.5% Nov 2026 Official leaderboard
10 GLM-5.3 61.4% Aug 2026 Official leaderboard
11 Muse Wonder 61.2% Oct 2026 Official leaderboard
12 Muse Ember 60.9% Nov 2026 Official leaderboard
13 DeepSeek V4.1 Flash 1017 60.9% Oct 2026 Official leaderboard
14 Gemini 3.8 Flash 60.5% Sep 2026 Official leaderboard
15 Grok 4.7 58.8% Sep 2026 Official leaderboard
16 Tencent Hy4 58% Nov 2026 Official leaderboard
17 Kimi K3 58% Jul 2026 Official leaderboard
18 MiMo-V2.6 Pro 57.9% Sep 2026 Official leaderboard
19 Tencent Hy4 preview 57.5% Aug 2026 Official leaderboard
20 Claude Haiku 5.5 57.5% Oct 2026 Official leaderboard
21 MiMo-V2.6 Flash 57.1% Sep 2026 Official leaderboard
22 DeepSeek V4.1 Flash 57.1% Sep 2026 Official leaderboard
23 Mistral Large 4 57% Oct 2026 Official leaderboard
24 Mistral Large 4 Preview 56.9% Oct 2026 Official leaderboard
25 Kimi K3.1 Pro 56% Oct 2026 Official leaderboard
26 MiniMax-M3.1-Flash 55.9% Oct 2026 Official leaderboard
27 GLM-5.3-Flash 55.8% Aug 2026 Official leaderboard
28 GPT-6.1 Sol 55.8% Sep 2026 Official leaderboard
29 GPT-6 Astra 55.6% Sep 2026 Official leaderboard
30 Step 5 Preview 55.4% Sep 2026 Official leaderboard
31 GPT-6.1 Terra 53.3% Oct 2026 Official leaderboard
32 GPT-6 Luna 53% Sep 2026 Official leaderboard
33 Muse Spark Open 53% Oct 2026 Official leaderboard
34 DeepSeek V4 Pro 0813 52.9% Aug 2026 Official leaderboard
35 MiniMax-M3 48.4% Jun 2026 Official leaderboard
36 Mercury 3 47.2% Oct 2026 Official leaderboard
37 Beam 1.0 46.5% Oct 2026 Official leaderboard
38 Nemotron 3.5 Super VL 41% Oct 2026 Official leaderboard
39 Inkling 38.3% Jul 2026 Official leaderboard
40 Inkling-Small 36.3% Jul 2026 Official leaderboard
#ModelScoreSettingDateSource
1 Gemini 4 Carbon 67% Oct 2026 Official leaderboard
2 Gemini 4 Hydrogen 66.1% Nov 2026 Official leaderboard
3 Gemini 4 Argon 66% Sep 2026 Official leaderboard
4 Claude Opus 5.5 65.3% Sep 2026 Official leaderboard
5 Muse Wonder Full 64.5% Oct 2026 Official leaderboard
6 Claude Sonnet 5.5 64.5% Sep 2026 Official leaderboard
7 Claude Fable 5.2 63.6% Nov 2026 Official leaderboard
8 Qwen3.8 Max (0902) 61.7% Sep 2026 Official leaderboard
9 Inkling 2 Small 61.5% Nov 2026 Lab model card
10 GLM-5.3 61.4% Aug 2026 Official leaderboard
11 Muse Wonder 61.2% Oct 2026 Official leaderboard
12 Muse Ember 60.9% Nov 2026 Official leaderboard
13 DeepSeek V4.1 Flash 1017 60.9% Oct 2026 Official leaderboard
14 Gemini 3.8 Flash 60.5% Sep 2026 Official leaderboard
15 Grok 4.7 58.8% Sep 2026 Official leaderboard
16 Tencent Hy4 58% Nov 2026 Official leaderboard
17 Kimi K3 58% Jul 2026 Official leaderboard
18 MiMo-V2.6 Pro 57.9% Sep 2026 Official leaderboard
19 Tencent Hy4 preview 57.5% Aug 2026 Official leaderboard
20 Claude Haiku 5.5 57.5% Oct 2026 Official leaderboard
21 MiMo-V2.6 Flash 57.1% Sep 2026 Official leaderboard
22 DeepSeek V4.1 Flash 57.1% Sep 2026 Official leaderboard
23 Mistral Large 4 57% Oct 2026 Official leaderboard
24 Mistral Large 4 Preview 56.9% Oct 2026 Official leaderboard
25 Kimi K3.1 Pro 56% Oct 2026 Official leaderboard
26 MiniMax-M3.1-Flash 55.9% Oct 2026 Official leaderboard
27 GLM-5.3-Flash 55.8% Aug 2026 Official leaderboard
28 GPT-6.1 Sol 55.8% Sep 2026 Official leaderboard
29 GPT-6 Astra 55.6% Sep 2026 Official leaderboard
30 Step 5 Preview 55.4% Sep 2026 Official leaderboard
31 GPT-6.1 Terra 53.3% Oct 2026 Official leaderboard
32 GPT-6 Luna 53% Sep 2026 Official leaderboard
33 Muse Spark Open 53% Oct 2026 Official leaderboard
34 DeepSeek V4 Pro 0813 52.9% Aug 2026 Official leaderboard
35 MiniMax-M3 48.4% Jun 2026 Official leaderboard
36 Mercury 3 47.2% Oct 2026 Official leaderboard
37 Beam 1.0 46.5% Oct 2026 Official leaderboard
38 Nemotron 3.5 Super VL 41% Oct 2026 Official leaderboard
39 Inkling 38.3% Jul 2026 Official leaderboard
40 Inkling-Small 36.3% Jul 2026 Official leaderboard

Labs report scores for 1 model in their cards. Switch the scores above to rank them together.

Accuracy against latency

The Vals scatter chart with its x axis switched from Cost to Latency: mean time per test against accuracy, on a log scale. The line joins the runs that no other run beats on both measures. We redrew this chart from the numbers on the official leaderboard.

40% 50% 60% 70% 80% 1001,00010K Latency (mean per test) (s), log scale, lower is better Accuracy MiniMax-M3: Accuracy 48.4%, latency (mean per test) 601 s Inkling: Accuracy 38.3%, latency (mean per test) 989 s Inkling-Small: Accuracy 36.3%, latency (mean per test) 624 s Kimi K3: Accuracy 58%, latency (mean per test) 1,344 s DeepSeek V4 Pro 0813: Accuracy 52.9%, latency (mean per test) 1,440 s GLM-5.3: Accuracy 61.4%, latency (mean per test) 2,144 s GLM-5.3-Flash: Accuracy 55.8%, latency (mean per test) 2,029 s Tencent Hy4 preview: Accuracy 57.5%, latency (mean per test) 1,935 s Qwen3.8 Max (0902): Accuracy 61.7%, latency (mean per test) 3,915 s Gemini 3.8 Flash: Accuracy 60.5%, latency (mean per test) 464 s GPT-6 Astra: Accuracy 55.6%, latency (mean per test) 978 s DeepSeek V4.1 Flash: Accuracy 57.1%, latency (mean per test) 697 s Step 5 Preview: Accuracy 55.4%, latency (mean per test) 1,482 s Grok 4.7: Accuracy 58.8%, latency (mean per test) 1,249 s MiMo-V2.6 Flash: Accuracy 57.1%, latency (mean per test) 1,166 s MiMo-V2.6 Pro: Accuracy 57.9%, latency (mean per test) 1,848 s Claude Opus 5.5: Accuracy 65.3%, latency (mean per test) 2,700 s GPT-6 Luna: Accuracy 53%, latency (mean per test) 1,321 s Claude Sonnet 5.5: Accuracy 64.5%, latency (mean per test) 2,611 s GPT-6.1 Sol: Accuracy 55.8%, latency (mean per test) 1,156 s Gemini 4 Argon: Accuracy 66%, latency (mean per test) 1,362 s Beam 1.0: Accuracy 46.5%, latency (mean per test) 449 s Mistral Large 4: Accuracy 57%, latency (mean per test) 933 s Mistral Large 4 Preview: Accuracy 56.9%, latency (mean per test) 3,110 s Claude Haiku 5.5: Accuracy 57.5%, latency (mean per test) 2,011 s Muse Wonder: Accuracy 61.2%, latency (mean per test) 612 s Muse Wonder Full: Accuracy 64.5%, latency (mean per test) 1,206 s Muse Spark Open: Accuracy 53%, latency (mean per test) 2,464 s Nemotron 3.5 Super VL: Accuracy 41%, latency (mean per test) 323 s DeepSeek V4.1 Flash 1017: Accuracy 60.9%, latency (mean per test) 1,382 s Mercury 3: Accuracy 47.2%, latency (mean per test) 344 s MiniMax-M3.1-Flash: Accuracy 55.9%, latency (mean per test) 1,572 s Kimi K3.1 Pro: Accuracy 56%, latency (mean per test) 2,300 s GPT-6.1 Terra: Accuracy 53.3%, latency (mean per test) 1,765 s Tencent Hy4: Accuracy 58%, latency (mean per test) 2,369 s Inkling 2 Small: Accuracy 61.5%, latency (mean per test) 515 s Gemini 4 Carbon: Accuracy 67%, latency (mean per test) 1,838 s Claude Fable 5.2: Accuracy 63.6%, latency (mean per test) 495 s Muse Ember: Accuracy 60.9%, latency (mean per test) 156 s Gemini 4 Hydrogen: Accuracy 66.1%, latency (mean per test) 945 s Gemini 4 CarbonGemini 4 HydrogenClaude Fable 5.2Muse Ember 40% 50% 60% 70% 80% 1001,00010K Latency (mean per test) (s), log scale, lower is better Accuracy MiniMax-M3: Accuracy 48.4%, latency (mean per test) 601 s Inkling: Accuracy 38.3%, latency (mean per test) 989 s Inkling-Small: Accuracy 36.3%, latency (mean per test) 624 s Kimi K3: Accuracy 58%, latency (mean per test) 1,344 s DeepSeek V4 Pro 0813: Accuracy 52.9%, latency (mean per test) 1,440 s GLM-5.3: Accuracy 61.4%, latency (mean per test) 2,144 s GLM-5.3-Flash: Accuracy 55.8%, latency (mean per test) 2,029 s Tencent Hy4 preview: Accuracy 57.5%, latency (mean per test) 1,935 s Qwen3.8 Max (0902): Accuracy 61.7%, latency (mean per test) 3,915 s Gemini 3.8 Flash: Accuracy 60.5%, latency (mean per test) 464 s GPT-6 Astra: Accuracy 55.6%, latency (mean per test) 978 s DeepSeek V4.1 Flash: Accuracy 57.1%, latency (mean per test) 697 s Step 5 Preview: Accuracy 55.4%, latency (mean per test) 1,482 s Grok 4.7: Accuracy 58.8%, latency (mean per test) 1,249 s MiMo-V2.6 Flash: Accuracy 57.1%, latency (mean per test) 1,166 s MiMo-V2.6 Pro: Accuracy 57.9%, latency (mean per test) 1,848 s Claude Opus 5.5: Accuracy 65.3%, latency (mean per test) 2,700 s GPT-6 Luna: Accuracy 53%, latency (mean per test) 1,321 s Claude Sonnet 5.5: Accuracy 64.5%, latency (mean per test) 2,611 s GPT-6.1 Sol: Accuracy 55.8%, latency (mean per test) 1,156 s Gemini 4 Argon: Accuracy 66%, latency (mean per test) 1,362 s Beam 1.0: Accuracy 46.5%, latency (mean per test) 449 s Mistral Large 4: Accuracy 57%, latency (mean per test) 933 s Mistral Large 4 Preview: Accuracy 56.9%, latency (mean per test) 3,110 s Claude Haiku 5.5: Accuracy 57.5%, latency (mean per test) 2,011 s Muse Wonder: Accuracy 61.2%, latency (mean per test) 612 s Muse Wonder Full: Accuracy 64.5%, latency (mean per test) 1,206 s Muse Spark Open: Accuracy 53%, latency (mean per test) 2,464 s Nemotron 3.5 Super VL: Accuracy 41%, latency (mean per test) 323 s DeepSeek V4.1 Flash 1017: Accuracy 60.9%, latency (mean per test) 1,382 s Mercury 3: Accuracy 47.2%, latency (mean per test) 344 s MiniMax-M3.1-Flash: Accuracy 55.9%, latency (mean per test) 1,572 s Kimi K3.1 Pro: Accuracy 56%, latency (mean per test) 2,300 s GPT-6.1 Terra: Accuracy 53.3%, latency (mean per test) 1,765 s Tencent Hy4: Accuracy 58%, latency (mean per test) 2,369 s Inkling 2 Small: Accuracy 61.5%, latency (mean per test) 515 s Gemini 4 Carbon: Accuracy 67%, latency (mean per test) 1,838 s Claude Fable 5.2: Accuracy 63.6%, latency (mean per test) 495 s Muse Ember: Accuracy 60.9%, latency (mean per test) 156 s Gemini 4 Hydrogen: Accuracy 66.1%, latency (mean per test) 945 s Gemini 4 CarbonGemini 4 HydrogenClaude Fable 5.2Muse Ember
ModelSettingAccuracyLatency (mean per test)
Gemini 4 Carbon 67.1% 1,838 s
Gemini 4 Hydrogen 66.1% 945.17 s
Gemini 4 Argon 66% 1,362 s
Claude Opus 5.5 65.3% 2,700 s
Muse Wonder Full 64.6% 1,206 s
Claude Sonnet 5.5 64.5% 2,611 s
Claude Fable 5.2 63.6% 495.48 s
Qwen3.8 Max (0902) 61.7% 3,915 s
Inkling 2 Small 61.5% 514.76 s
GLM-5.3 61.4% 2,144 s
Muse Wonder 61.2% 612.47 s
Muse Ember 60.9% 156.44 s
DeepSeek V4.1 Flash 1017 60.9% 1,382 s
Gemini 3.8 Flash 60.5% 463.81 s
Grok 4.7 58.8% 1,249 s
Tencent Hy4 58.1% 2,369 s
Kimi K3 58% 1,344 s
MiMo-V2.6 Pro 57.9% 1,848 s
Tencent Hy4 preview 57.5% 1,935 s
Claude Haiku 5.5 57.5% 2,011 s
MiMo-V2.6 Flash 57.2% 1,166 s
DeepSeek V4.1 Flash 57.1% 696.59 s
Mistral Large 4 57% 932.62 s
Mistral Large 4 Preview 56.9% 3,110 s
Kimi K3.1 Pro 56% 2,300 s
MiniMax-M3.1-Flash 55.9% 1,572 s
GLM-5.3-Flash 55.8% 2,029 s
GPT-6.1 Sol 55.8% 1,156 s
GPT-6 Astra 55.6% 977.87 s
Step 5 Preview 55.4% 1,482 s
GPT-6.1 Terra 53.3% 1,765 s
GPT-6 Luna 53.1% 1,321 s
Muse Spark Open 53% 2,464 s
DeepSeek V4 Pro 0813 52.9% 1,440 s
MiniMax-M3 48.4% 600.9 s
Mercury 3 47.2% 344.44 s
Beam 1.0 46.5% 449.23 s
Nemotron 3.5 Super VL 41% 322.96 s
Inkling 38.3% 988.71 s
Inkling-Small 36.3% 624.36 s

The best score, 67%, costs $6.46 a task

Runs on the cost frontier score higher than any run that's cheaper. Muse Wonder Full comes within 2.5 points of the best for $3.14 a task, about half the cost. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

Meta Google DeepMind Thinking Machines Cost frontier One model at each effort level
0% 20% 40% 60% 80% $0.01 $0.10 $1.00 $10 $100 Cost per task, log scale Gemini 4 Carbon Inkling 2 Small Muse Ember 0% 20% 40% 60% 80% $0.01 $0.10 $1.00 $10 $100 Cost per task, log scale Gemini 4 Carbon Inkling 2 Small

Source: Official Vals Finance Index leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Gemini 4 Carbon 67% $6.46 160,960
Gemini 4 Hydrogen 66.1% $3.80 256,546
Gemini 4 Argon 66% $6.53 95,106
Claude Opus 5.5 65.3% $11 312,134
Muse Wonder Full 64.5% $3.14 57,946
Claude Sonnet 5.5 64.5% $7.40 395,199
Claude Fable 5.2 63.6% $4.94 48,712
Qwen3.8 Max (0902) 61.7% $3.83 197,086
Inkling 2 Small 61.5% $0.28 87,270
GLM-5.3 61.4% $2.21 169,379
Muse Wonder 61.2% $3.37 53,074
Muse Ember 60.9% $0.06 44,388
DeepSeek V4.1 Flash 1017 60.9% $0.42 215,664
Gemini 3.8 Flash 60.5% $3.70 102,660
Grok 4.7 58.8% $3.51 86,233
Tencent Hy4 58% $1.20 251,853
Kimi K3 58% $2.58 69,391
MiMo-V2.6 Pro 57.9% $0.22 87,303
Tencent Hy4 preview 57.5% $0.68 121,099
Claude Haiku 5.5 57.5% $1.39 482,804
MiMo-V2.6 Flash 57.1% $0.07 84,370
DeepSeek V4.1 Flash 57.1% $0.25 132,306
Mistral Large 4 57% $0.56 67,899
Mistral Large 4 Preview 56.9% $1.46 125,008
Kimi K3.1 Pro 56% $2.51 117,528
MiniMax-M3.1-Flash 55.9% $0.25 208,140
GLM-5.3-Flash 55.8% $0.55 159,905
GPT-6.1 Sol 55.8% $1.36 58,087
GPT-6 Astra 55.6% $6.48 55,355
Step 5 Preview 55.4% $0.82 91,378
GPT-6.1 Terra 53.3% $1.55 150,291
GPT-6 Luna 53% $0.15 141,399
Muse Spark Open 53% $3.00 44,557
DeepSeek V4 Pro 0813 52.9% $0.88 106,288
MiniMax-M3 48.4% $1.22 63,084
Mercury 3 47.2% $0.10 78,420
Beam 1.0 46.5% 39,315
Nemotron 3.5 Super VL 41% $0.48 93,952
Inkling 38.3% $0.97 63,992
Inkling-Small 36.3% $0.18 51,276

2 models have held number one on the official leaderboard since its release in October 2026 2 models have held number one with claims included since its release in October 2026

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

Google DeepMind
Own leaderboard 2 leaders
Gemini 4 Carbon 67%

Source: Official Vals Finance Index 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 Reported by the lab in its model card Best so far
0% 20% 40% 60% 80% 100% 2026 2027 Near the ceiling (85%) Gemini 4 Carbon Qwen3.8 Max (0902) MiniMax-M3 0% 20% 40% 60% 80% 100% 2026 2027 Near the ceiling (85%) Gemini 4 Carbon Kimi K3

Sources: Official Vals Finance Index leaderboard and labs’ model cards

One lab put it in a launch post in the last six months

Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.

Thinking Machines
DateLaunch postScoreAs the post writes it
Nov 5, 2026 Inkling 2 Small 61.5 Vals Finance Index

One lab claim next to the leaderboard

Each score a lab reported for its own model, beside the same model's best score on the leaderboard. Gaps mostly come from different harnesses, effort levels and task subsets. 1 of them match, which usually means one side copied the other. Every benchmark.

Model Lab says Board says Gap
Inkling 2 Small 61.5% Vals Finance Index 61.5% Leaderboard Same

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%