Benchmark Almanac

EKI Eesti mäng

Gemini 4 Argon 90% leads since September 2026
Not in model cards

Models choose answers to 1,000 questions from an Estonian quiz book. The questions cover subjects including history, culture, and geography.

The institute reports accuracy for 101 models. Its 1,000 questions come from Tiit Kuningas’s book. TalTech’s separate Trivia-et uses board game cards.

Released
June 2026
Kind
Knowledge in many languages
Level
Everyday
Scored
Checked automatically
Answers
Multiple choice
Measures
Capability
Language
Estonian
In and out
Text → Text
Interaction
Single turn
Items from
Real work and use
Refresh
One fixed set
Run by
A host
Strongest model tested
Claude Fable 5.2 #2 on the capability index
Last model added
November 2026
Scores refreshed
November 20th, 2026
101
Entries on its own leaderboard, 96 models we track

Gemini 4 Argon leads with 90% Gemini 4 Argon leads with 90%

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. 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.

101 entries. Scroll sideways for the rest.

  1. 90.0 Gemini 4 Argon
  2. 85.0 Gemini 3.1 Pro
  3. 82.8 Gemini 3.8 Flash
  4. 81.9 Gemini 3.7 Flash
  5. 81.1 GPT-6 Astra
  6. 80.9 Gemini 3.6 Flash
  7. 80.5 Claude Fable 5.2
  8. 80.0 Gemini 3.5 Flash
  9. 78.0 GPT-5.6 Sol
  10. 76.6 GPT-5.5
  11. 76.4 GPT-6 Sol
  12. 76.1 Gemini 3 Flash
  13. 71.9 Claude Fable 5
  14. 71.5 Gemini 2.5 Pro
  15. 70.8 Claude Fable 5.1
  16. 69.2 GPT-6.1 Terra
  17. 67.9 Muse Spark Open
  18. 66.9 Muse Spark 1.2
  19. 66.7 GPT-6 Luna
  20. 64.9 Qwen3.7-Max
  21. 64.8 Claude Opus 5
  22. 64.4 Claude Opus 5.5
  23. 64.2 Grok 4.6
  24. 63.6 Muse Spark 1.3
  25. 63.4 Grok 4.7
  26. 63.2 Grok 4.5
  27. 62.9 GPT-5.6 Terra
  28. 62.7 GPT-5
  29. 62.4 DeepSeek-V4-Pro
  30. 61.3 GPT-5.6 Luna
  31. 58.2 GPT-5.2
  32. 57.6 DeepSeek V4.1 Flash
  33. 57.4 Claude Opus 4.7
  34. 56.9 Tencent Hy4 preview
  35. 55.8 Claude Opus 4.8
  36. 55.7 Kimi K3
  37. 54.4 Claude Opus 4.6
  38. 54.2 Claude Opus 4.5
  39. 54.0 GPT-4o
  40. 54.0 GLM-5.3-Flash
  41. 53.6 Gemini 2.5 Flash
  42. 53.2 GPT-4.1
  43. 53.0 GLM-5.2
  44. 52.8 Qwen 3.8 Max
  45. 52.4 GLM-5.3
  46. 52.1 GPT-5.4
  47. 51.6 Claude Opus 4
  48. 50.8 GPT-5.4 Mini
  49. 50.7 Claude 3.7 Sonnet
  50. 50.7 Qwen 3.6 Plus
  51. 50.3 GLM-5
  52. 50.1 Claude Sonnet 5.5
  53. 49.9 Grok 3
  54. 48.6 Gemini 2.0 Flash
  55. 47.8 MiMo-V2.6 Pro
  56. 47.4 Claude Sonnet 5
  57. 47.3 Mistral Large 2512
  58. 46.8 MiniMax-M3
  59. 46.5 Nemotron 3 Ultra
  60. 46.4 GPT-4
  61. 45.7 Claude Sonnet 4.6
  62. 45.3 Mistral Large 4
  63. 44.6 Llama 4 Maverick
  64. 42.7 Grok-3 mini
  65. 42.3 MiMo-V2.6 Flash
  66. 42.1 Claude Sonnet 4.5
  67. 42.0 Kimi K2
  68. 42.0 Mistral Medium 3.5
  69. 41.4 Qwen3.8 Flash
  70. 40.9 Grok 4 Fast
  71. 40.4 Claude Haiku 5.5
  72. 40.0 Muse Glimmer
  73. 39.5 Grok 4.20
  74. 39.4 Claude Sonnet 4
  75. 39.2 Claude 3.5 Haiku
  76. 39.2 DeepSeek-V3
  77. 38.4 DeepSeek-V4-Flash
  78. 37.8 DeepSeek-V3.2
  79. 37.3 Step 3.7 Flash
  80. 37.2 Mistral Small 4
  81. 36.5 Llama 3.1 70B Instruct
  82. 35.7 Llama 3.3 70B Instruct
  83. 35.3 Nemotron 3 Super 120B
  84. 35.3 GPT-3.5 Turbo
  85. 35.1 GPT-4o mini
  86. 35.0 Qwen3.6 27B
  87. 34.9 Claude Haiku 4.5
  88. 34.7 Gemma 4 31B IT
  89. 34.6 Mistral Large 2411
  90. 34.6 Gemma 3 27B
  91. 34.5 Laguna S 2.1
  92. 32.6 Claude 3 Haiku
  93. 32.1 Qwen 3.8 27B
  94. 32.0 Gemma 2 27B
  95. 31.9 Command R+ (08-2024)
  96. 31.9 Qwen3.5-27B
  97. 25.7 Step 3.5 Flash
  98. 23.0 MiniMax-M2.7
  99. 22.5 MiMo-V2-Pro
  100. 21.0 GLM-4.7
  101. 20.2 gpt-oss-120b
  1. 1 Gemini 4 Argon 90%
  2. 2 Gemini 3.1 Pro 85%
  3. 3 Gemini 3.8 Flash 82.8%
  4. 4 Gemini 3.7 Flash 81.9%
  5. 5 GPT-6 Astra 81.1%
  6. 6 Gemini 3.6 Flash 80.9%
  7. 7 Claude Fable 5.2 80.5%
  8. 8 Gemini 3.5 Flash 80%
  9. 9 GPT-5.6 Sol 78%
  10. 10 GPT-5.5 76.6%
  11. 11 GPT-6 Sol 76.4%
  12. 12 Gemini 3 Flash 76.1%
  13. 13 Claude Fable 5 71.9%
  14. 14 Gemini 2.5 Pro Jun 2025 71.5%
  15. 15 Claude Fable 5.1 70.8%
  16. 16 GPT-6.1 Terra 69.2%
  17. 17 Muse Spark Open 67.9%
  18. 18 Muse Spark 1.2 66.9%
  19. 19 GPT-6 Luna 66.7%
  20. 20 Qwen3.7-Max 64.9%
  21. 21 Claude Opus 5 64.8%
  22. 22 Claude Opus 5.5 64.4%
  23. 23 Grok 4.6 64.2%
  24. 24 Muse Spark 1.3 63.6%
  25. 25 Grok 4.7 63.4%
  26. 26 Grok 4.5 63.2%
  27. 27 GPT-5.6 Terra 62.9%
  28. 28 GPT-5 62.7%
  29. 29 DeepSeek-V4-Pro 62.4%
  30. 30 GPT-5.6 Luna 61.3%
  31. 31 GPT-5.2 58.2%
  32. 32 DeepSeek V4.1 Flash 57.6%
  33. 33 Claude Opus 4.7 57.4%
  34. 34 Tencent Hy4 preview 56.9%
  35. 35 Claude Opus 4.8 55.8%
  36. 36 Kimi K3 55.7%
  37. 37 Claude Opus 4.6 54.4%
  38. 38 Claude Opus 4.5 54.2%
  39. 39 GPT-4o Mar 2025 54%
  40. 40 GLM-5.3-Flash 54%
  41. 41 Gemini 2.5 Flash Sep 2025 53.6%
  42. 42 GPT-4.1 53.2%
  43. 43 GLM-5.2 53%
  44. 44 Qwen 3.8 Max 52.8%
  45. 45 GLM-5.3 52.4%
  46. 46 GPT-5.4 52.1%
  47. 47 Claude Opus 4 51.6%
  48. 48 GPT-5.4 Mini 50.8%
  49. 49 Claude 3.7 Sonnet 50.7%
  50. 50 Qwen 3.6 Plus 50.7%
  51. 51 GLM-5 50.3%
  52. 52 Claude Sonnet 5.5 50.1%
  53. 53 Grok 3 49.9%
  54. 54 Gemini 2.0 Flash Feb 2025 48.6%
  55. 55 MiMo-V2.6 Pro 47.8%
  56. 56 Claude Sonnet 5 47.4%
  57. 57 Mistral Large 2512 47.3%
  58. 58 MiniMax-M3 46.8%
  59. 59 Nemotron 3 Ultra 46.5%
  60. 60 GPT-4 Jun 2023 46.4%
  61. 61 Claude Sonnet 4.6 45.7%
  62. 62 Mistral Large 4 45.3%
  63. 63 Llama 4 Maverick 44.6%
  64. 64 Grok-3 mini 42.7%
  65. 65 MiMo-V2.6 Flash 42.3%
  66. 66 Claude Sonnet 4.5 42.1%
  67. 67 Kimi K2 Sep 2025 42%
  68. 68 Mistral Medium 3.5 42%
  69. 69 Qwen3.8 Flash 41.4%
  70. 70 Grok 4 Fast 40.9%
  71. 71 Claude Haiku 5.5 40.4%
  72. 72 Muse Glimmer 40%
  73. 73 Grok 4.20 39.5%
  74. 74 Claude Sonnet 4 39.4%
  75. 75 Claude 3.5 Haiku 39.2%
  76. 76 DeepSeek-V3 Mar 2025 39.2%
  77. 77 DeepSeek-V4-Flash 38.4%
  78. 78 DeepSeek-V3.2 37.8%
  79. 79 Step 3.7 Flash 37.3%
  80. 80 Mistral Small 4 37.2%
  81. 81 Llama 3.1 70B Instruct 36.5%
  82. 82 Llama 3.3 70B Instruct 35.7%
  83. 83 Nemotron 3 Super 120B 35.3%
  84. 84 GPT-3.5 Turbo Jan 2024 35.3%
  85. 85 GPT-4o mini 35.1%
  86. 86 Qwen3.6 27B 35%
  87. 87 Claude Haiku 4.5 34.9%
  88. 88 Gemma 4 31B IT 34.7%
  89. 89 Mistral Large 2411 34.6%
  90. 90 Gemma 3 27B 34.6%
  91. 91 Laguna S 2.1 34.5%
  92. 92 Claude 3 Haiku 32.6%
  93. 93 Qwen 3.8 27B 32.1%
  94. 94 Gemma 2 27B 32%
  95. 95 Command R+ (08-2024) 31.9%
  96. 96 Qwen3.5-27B 31.9%
  97. 97 Step 3.5 Flash 25.7%
  98. 98 MiniMax-M2.7 23%
  99. 99 MiMo-V2-Pro 22.5%
  100. 100 GLM-4.7 21%
  101. 101 gpt-oss-120b 20.2%

Source: Official EKI Eesti mäng leaderboard

96 models. Scroll sideways for the rest.

  1. 90.0 Gemini 4 Argon
  2. 85.0 Gemini 3.1 Pro
  3. 82.8 Gemini 3.8 Flash
  4. 81.9 Gemini 3.7 Flash
  5. 81.1 GPT-6 Astra
  6. 80.9 Gemini 3.6 Flash
  7. 80.5 Claude Fable 5.2
  8. 80.0 Gemini 3.5 Flash
  9. 78.0 GPT-5.6 Sol
  10. 76.6 GPT-5.5
  11. 76.4 GPT-6 Sol
  12. 76.1 Gemini 3 Flash
  13. 71.9 Claude Fable 5
  14. 71.5 Gemini 2.5 Pro
  15. 70.8 Claude Fable 5.1
  16. 69.2 GPT-6.1 Terra
  17. 67.9 Muse Spark Open
  18. 66.9 Muse Spark 1.2
  19. 66.7 GPT-6 Luna
  20. 64.9 Qwen3.7-Max
  21. 64.8 Claude Opus 5
  22. 64.4 Claude Opus 5.5
  23. 64.2 Grok 4.6
  24. 63.6 Muse Spark 1.3
  25. 63.4 Grok 4.7
  26. 63.2 Grok 4.5
  27. 62.9 GPT-5.6 Terra
  28. 62.7 GPT-5
  29. 62.4 DeepSeek-V4-Pro
  30. 61.3 GPT-5.6 Luna
  31. 58.2 GPT-5.2
  32. 57.6 DeepSeek V4.1 Flash
  33. 57.4 Claude Opus 4.7
  34. 56.9 Tencent Hy4 preview
  35. 55.8 Claude Opus 4.8
  36. 55.7 Kimi K3
  37. 54.4 Claude Opus 4.6
  38. 54.2 Claude Opus 4.5
  39. 54.0 GPT-4o
  40. 54.0 GLM-5.3-Flash
  41. 53.6 Gemini 2.5 Flash
  42. 53.2 GPT-4.1
  43. 53.0 GLM-5.2
  44. 52.8 Qwen 3.8 Max
  45. 52.4 GLM-5.3
  46. 52.1 GPT-5.4
  47. 51.6 Claude Opus 4
  48. 50.8 GPT-5.4 Mini
  49. 50.7 Claude 3.7 Sonnet
  50. 50.7 Qwen 3.6 Plus
  51. 50.3 GLM-5
  52. 50.1 Claude Sonnet 5.5
  53. 49.9 Grok 3
  54. 48.6 Gemini 2.0 Flash
  55. 47.8 MiMo-V2.6 Pro
  56. 47.4 Claude Sonnet 5
  57. 46.8 MiniMax-M3
  58. 46.5 Nemotron 3 Ultra
  59. 46.4 GPT-4
  60. 45.7 Claude Sonnet 4.6
  61. 45.3 Mistral Large 4
  62. 44.6 Llama 4 Maverick
  63. 42.7 Grok-3 mini
  64. 42.3 MiMo-V2.6 Flash
  65. 42.1 Claude Sonnet 4.5
  66. 42.0 Kimi K2
  67. 42.0 Mistral Medium 3.5
  68. 41.4 Qwen3.8 Flash
  69. 40.9 Grok 4 Fast
  70. 40.4 Claude Haiku 5.5
  71. 40.0 Muse Glimmer
  72. 39.5 Grok 4.20
  73. 39.4 Claude Sonnet 4
  74. 39.2 Claude 3.5 Haiku
  75. 39.2 DeepSeek-V3
  76. 38.4 DeepSeek-V4-Flash
  77. 37.8 DeepSeek-V3.2
  78. 37.3 Step 3.7 Flash
  79. 37.2 Mistral Small 4
  80. 35.7 Llama 3.3 70B Instruct
  81. 35.3 GPT-3.5 Turbo
  82. 35.1 GPT-4o mini
  83. 35.0 Qwen3.6 27B
  84. 34.9 Claude Haiku 4.5
  85. 34.7 Gemma 4 31B IT
  86. 34.6 Gemma 3 27B
  87. 34.5 Laguna S 2.1
  88. 32.6 Claude 3 Haiku
  89. 32.1 Qwen 3.8 27B
  90. 32.0 Gemma 2 27B
  91. 31.9 Qwen3.5-27B
  92. 25.7 Step 3.5 Flash
  93. 23.0 MiniMax-M2.7
  94. 22.5 MiMo-V2-Pro
  95. 21.0 GLM-4.7
  96. 20.2 gpt-oss-120b
  1. 1 Gemini 4 Argon 90%
  2. 2 Gemini 3.1 Pro 85%
  3. 3 Gemini 3.8 Flash 82.8%
  4. 4 Gemini 3.7 Flash 81.9%
  5. 5 GPT-6 Astra 81.1%
  6. 6 Gemini 3.6 Flash 80.9%
  7. 7 Claude Fable 5.2 80.5%
  8. 8 Gemini 3.5 Flash 80%
  9. 9 GPT-5.6 Sol 78%
  10. 10 GPT-5.5 76.6%
  11. 11 GPT-6 Sol 76.4%
  12. 12 Gemini 3 Flash 76.1%
  13. 13 Claude Fable 5 71.9%
  14. 14 Gemini 2.5 Pro Jun 2025 71.5%
  15. 15 Claude Fable 5.1 70.8%
  16. 16 GPT-6.1 Terra 69.2%
  17. 17 Muse Spark Open 67.9%
  18. 18 Muse Spark 1.2 66.9%
  19. 19 GPT-6 Luna 66.7%
  20. 20 Qwen3.7-Max 64.9%
  21. 21 Claude Opus 5 64.8%
  22. 22 Claude Opus 5.5 64.4%
  23. 23 Grok 4.6 64.2%
  24. 24 Muse Spark 1.3 63.6%
  25. 25 Grok 4.7 63.4%
  26. 26 Grok 4.5 63.2%
  27. 27 GPT-5.6 Terra 62.9%
  28. 28 GPT-5 62.7%
  29. 29 DeepSeek-V4-Pro 62.4%
  30. 30 GPT-5.6 Luna 61.3%
  31. 31 GPT-5.2 58.2%
  32. 32 DeepSeek V4.1 Flash 57.6%
  33. 33 Claude Opus 4.7 57.4%
  34. 34 Tencent Hy4 preview 56.9%
  35. 35 Claude Opus 4.8 55.8%
  36. 36 Kimi K3 55.7%
  37. 37 Claude Opus 4.6 54.4%
  38. 38 Claude Opus 4.5 54.2%
  39. 39 GPT-4o Mar 2025 54%
  40. 40 GLM-5.3-Flash 54%
  41. 41 Gemini 2.5 Flash Sep 2025 53.6%
  42. 42 GPT-4.1 53.2%
  43. 43 GLM-5.2 53%
  44. 44 Qwen 3.8 Max 52.8%
  45. 45 GLM-5.3 52.4%
  46. 46 GPT-5.4 52.1%
  47. 47 Claude Opus 4 51.6%
  48. 48 GPT-5.4 Mini 50.8%
  49. 49 Claude 3.7 Sonnet 50.7%
  50. 50 Qwen 3.6 Plus 50.7%
  51. 51 GLM-5 50.3%
  52. 52 Claude Sonnet 5.5 50.1%
  53. 53 Grok 3 49.9%
  54. 54 Gemini 2.0 Flash Feb 2025 48.6%
  55. 55 MiMo-V2.6 Pro 47.8%
  56. 56 Claude Sonnet 5 47.4%
  57. 57 MiniMax-M3 46.8%
  58. 58 Nemotron 3 Ultra 46.5%
  59. 59 GPT-4 Jun 2023 46.4%
  60. 60 Claude Sonnet 4.6 45.7%
  61. 61 Mistral Large 4 45.3%
  62. 62 Llama 4 Maverick 44.6%
  63. 63 Grok-3 mini 42.7%
  64. 64 MiMo-V2.6 Flash 42.3%
  65. 65 Claude Sonnet 4.5 42.1%
  66. 66 Kimi K2 Sep 2025 42%
  67. 67 Mistral Medium 3.5 42%
  68. 68 Qwen3.8 Flash 41.4%
  69. 69 Grok 4 Fast 40.9%
  70. 70 Claude Haiku 5.5 40.4%
  71. 71 Muse Glimmer 40%
  72. 72 Grok 4.20 39.5%
  73. 73 Claude Sonnet 4 39.4%
  74. 74 Claude 3.5 Haiku 39.2%
  75. 75 DeepSeek-V3 Mar 2025 39.2%
  76. 76 DeepSeek-V4-Flash 38.4%
  77. 77 DeepSeek-V3.2 37.8%
  78. 78 Step 3.7 Flash 37.3%
  79. 79 Mistral Small 4 37.2%
  80. 80 Llama 3.3 70B Instruct 35.7%
  81. 81 GPT-3.5 Turbo Jan 2024 35.3%
  82. 82 GPT-4o mini 35.1%
  83. 83 Qwen3.6 27B 35%
  84. 84 Claude Haiku 4.5 34.9%
  85. 85 Gemma 4 31B IT 34.7%
  86. 86 Gemma 3 27B 34.6%
  87. 87 Laguna S 2.1 34.5%
  88. 88 Claude 3 Haiku 32.6%
  89. 89 Qwen 3.8 27B 32.1%
  90. 90 Gemma 2 27B 32%
  91. 91 Qwen3.5-27B 31.9%
  92. 92 Step 3.5 Flash 25.7%
  93. 93 MiniMax-M2.7 23%
  94. 94 MiMo-V2-Pro 22.5%
  95. 95 GLM-4.7 21%
  96. 96 gpt-oss-120b 20.2%

Source: Official EKI Eesti mäng leaderboard

#EntryScoreSettingDateSource
1 Gemini 4 Argon 90% Sep 2026 Official leaderboard
2 Gemini 3.1 Pro 85% Feb 2026 Official leaderboard
3 Gemini 3.8 Flash 82.8% Sep 2026 Official leaderboard
4 Gemini 3.7 Flash 81.9% Aug 2026 Official leaderboard
5 GPT-6 Astra 81.1% Sep 2026 Official leaderboard
6 Gemini 3.6 Flash 80.9% Jul 2026 Official leaderboard
7 Claude Fable 5.2 80.5% Nov 2026 Official leaderboard
8 Gemini 3.5 Flash 80% May 2026 Official leaderboard
9 GPT-5.6 Sol 78% Jul 2026 Official leaderboard
10 GPT-5.5 76.6% Apr 2026 Official leaderboard
11 GPT-6 Sol 76.4% Sep 2026 Official leaderboard
12 Gemini 3 Flash 76.1% Dec 2025 Official leaderboard
13 Claude Fable 5 71.9% Jun 2026 Official leaderboard
14 Gemini 2.5 Pro Jun 2025 71.5% Jun 2025 Official leaderboard
15 Claude Fable 5.1 70.8% Sep 2026 Official leaderboard
16 GPT-6.1 Terra 69.2% Oct 2026 Official leaderboard
17 Muse Spark Open 67.9% Oct 2026 Official leaderboard
18 Muse Spark 1.2 66.9% Aug 2026 Official leaderboard
19 GPT-6 Luna 66.7% Sep 2026 Official leaderboard
20 Qwen3.7-Max 64.9% May 2026 Official leaderboard
21 Claude Opus 5 64.8% Jul 2026 Official leaderboard
22 Claude Opus 5.5 64.4% Sep 2026 Official leaderboard
23 Grok 4.6 64.2% Aug 2026 Official leaderboard
24 Muse Spark 1.3 63.6% Sep 2026 Official leaderboard
25 Grok 4.7 63.4% Sep 2026 Official leaderboard
26 Grok 4.5 63.2% Jul 2026 Official leaderboard
27 GPT-5.6 Terra 62.9% Jul 2026 Official leaderboard
28 GPT-5 62.7% Aug 2025 Official leaderboard
29 DeepSeek-V4-Pro 62.4% Apr 2026 Official leaderboard
30 GPT-5.6 Luna 61.3% Jul 2026 Official leaderboard
31 GPT-5.2 58.2% Dec 2025 Official leaderboard
32 DeepSeek V4.1 Flash 57.6% Sep 2026 Official leaderboard
33 Claude Opus 4.7 57.4% Apr 2026 Official leaderboard
34 Tencent Hy4 preview 56.9% Aug 2026 Official leaderboard
35 Claude Opus 4.8 55.8% May 2026 Official leaderboard
36 Kimi K3 55.7% Jul 2026 Official leaderboard
37 Claude Opus 4.6 54.4% Feb 2026 Official leaderboard
38 Claude Opus 4.5 54.2% Nov 2025 Official leaderboard
39 GPT-4o Mar 2025 54% Mar 2025 Official leaderboard
40 GLM-5.3-Flash 54% Aug 2026 Official leaderboard
41 Gemini 2.5 Flash Sep 2025 53.6% Sep 2025 Official leaderboard
42 GPT-4.1 53.2% Apr 2025 Official leaderboard
43 GLM-5.2 53% Jun 2026 Official leaderboard
44 Qwen 3.8 Max 52.8% Aug 2026 Official leaderboard
45 GLM-5.3 52.4% Aug 2026 Official leaderboard
46 GPT-5.4 52.1% Mar 2026 Official leaderboard
47 Claude Opus 4 51.6% May 2025 Official leaderboard
48 GPT-5.4 Mini 50.8% Mar 2026 Official leaderboard
49 Claude 3.7 Sonnet 50.7% Feb 2025 Official leaderboard
50 Qwen 3.6 Plus 50.7% Mar 2026 Official leaderboard
51 GLM-5 50.3% Feb 2026 Official leaderboard
52 Claude Sonnet 5.5 50.1% Sep 2026 Official leaderboard
53 Grok 3 49.9% Apr 2025 Official leaderboard
54 Gemini 2.0 Flash Feb 2025 48.6% Feb 2025 Official leaderboard
55 MiMo-V2.6 Pro 47.8% Sep 2026 Official leaderboard
56 Claude Sonnet 5 47.4% Jun 2026 Official leaderboard
57 Mistral Large 2512 open weights 47.3% Official leaderboard
58 MiniMax-M3 46.8% Jun 2026 Official leaderboard
59 Nemotron 3 Ultra 46.5% Jun 2026 Official leaderboard
60 GPT-4 Jun 2023 46.4% Jun 2023 Official leaderboard
61 Claude Sonnet 4.6 45.7% Feb 2026 Official leaderboard
62 Mistral Large 4 45.3% Oct 2026 Official leaderboard
63 Llama 4 Maverick 44.6% Apr 2025 Official leaderboard
64 Grok-3 mini 42.7% Apr 2025 Official leaderboard
65 MiMo-V2.6 Flash 42.3% Sep 2026 Official leaderboard
66 Claude Sonnet 4.5 42.1% Sep 2025 Official leaderboard
67 Kimi K2 Sep 2025 42% Sep 2025 Official leaderboard
68 Mistral Medium 3.5 42% Apr 2026 Official leaderboard
69 Qwen3.8 Flash 41.4% Aug 2026 Official leaderboard
70 Grok 4 Fast 40.9% Sep 2025 Official leaderboard
71 Claude Haiku 5.5 40.4% Oct 2026 Official leaderboard
72 Muse Glimmer 40% Aug 2026 Official leaderboard
73 Grok 4.20 39.5% Feb 2026 Official leaderboard
74 Claude Sonnet 4 39.4% May 2025 Official leaderboard
75 Claude 3.5 Haiku 39.2% Oct 2024 Official leaderboard
76 DeepSeek-V3 Mar 2025 39.2% Mar 2025 Official leaderboard
77 DeepSeek-V4-Flash 38.4% Apr 2026 Official leaderboard
78 DeepSeek-V3.2 37.8% Dec 2025 Official leaderboard
79 Step 3.7 Flash 37.3% May 2026 Official leaderboard
80 Mistral Small 4 37.2% Mar 2026 Official leaderboard
81 Llama 3.1 70B Instruct open weights 36.5% Official leaderboard
82 Llama 3.3 70B Instruct 35.7% Dec 2024 Official leaderboard
83 Nemotron 3 Super 120B open weights 35.3% Official leaderboard
84 GPT-3.5 Turbo Jan 2024 35.3% Jan 2024 Official leaderboard
85 GPT-4o mini 35.1% Jul 2024 Official leaderboard
86 Qwen3.6 27B 35% Apr 2026 Official leaderboard
87 Claude Haiku 4.5 34.9% Oct 2025 Official leaderboard
88 Gemma 4 31B IT 34.7% Apr 2026 Official leaderboard
89 Mistral Large 2411 open weights 34.6% Official leaderboard
90 Gemma 3 27B 34.6% Mar 2025 Official leaderboard
91 Laguna S 2.1 34.5% Jul 2026 Official leaderboard
92 Claude 3 Haiku 32.6% Mar 2024 Official leaderboard
93 Qwen 3.8 27B 32.1% Aug 2026 Official leaderboard
94 Gemma 2 27B 32% Jun 2024 Official leaderboard
95 Command R+ (08-2024) 31.9% Official leaderboard
96 Qwen3.5-27B 31.9% Feb 2026 Official leaderboard
97 Step 3.5 Flash 25.7% Feb 2026 Official leaderboard
98 MiniMax-M2.7 23% Mar 2026 Official leaderboard
99 MiMo-V2-Pro 22.5% Mar 2026 Official leaderboard
100 GLM-4.7 21% Dec 2025 Official leaderboard
101 gpt-oss-120b 20.2% Aug 2025 Official leaderboard
#ModelScoreSettingDateSource
1 Gemini 4 Argon 90% Sep 2026 Official leaderboard
2 Gemini 3.1 Pro 85% Feb 2026 Official leaderboard
3 Gemini 3.8 Flash 82.8% Sep 2026 Official leaderboard
4 Gemini 3.7 Flash 81.9% Aug 2026 Official leaderboard
5 GPT-6 Astra 81.1% Sep 2026 Official leaderboard
6 Gemini 3.6 Flash 80.9% Jul 2026 Official leaderboard
7 Claude Fable 5.2 80.5% Nov 2026 Official leaderboard
8 Gemini 3.5 Flash 80% May 2026 Official leaderboard
9 GPT-5.6 Sol 78% Jul 2026 Official leaderboard
10 GPT-5.5 76.6% Apr 2026 Official leaderboard
11 GPT-6 Sol 76.4% Sep 2026 Official leaderboard
12 Gemini 3 Flash 76.1% Dec 2025 Official leaderboard
13 Claude Fable 5 71.9% Jun 2026 Official leaderboard
14 Gemini 2.5 Pro Jun 2025 71.5% Jun 2025 Official leaderboard
15 Claude Fable 5.1 70.8% Sep 2026 Official leaderboard
16 GPT-6.1 Terra 69.2% Oct 2026 Official leaderboard
17 Muse Spark Open 67.9% Oct 2026 Official leaderboard
18 Muse Spark 1.2 66.9% Aug 2026 Official leaderboard
19 GPT-6 Luna 66.7% Sep 2026 Official leaderboard
20 Qwen3.7-Max 64.9% May 2026 Official leaderboard
21 Claude Opus 5 64.8% Jul 2026 Official leaderboard
22 Claude Opus 5.5 64.4% Sep 2026 Official leaderboard
23 Grok 4.6 64.2% Aug 2026 Official leaderboard
24 Muse Spark 1.3 63.6% Sep 2026 Official leaderboard
25 Grok 4.7 63.4% Sep 2026 Official leaderboard
26 Grok 4.5 63.2% Jul 2026 Official leaderboard
27 GPT-5.6 Terra 62.9% Jul 2026 Official leaderboard
28 GPT-5 62.7% Aug 2025 Official leaderboard
29 DeepSeek-V4-Pro 62.4% Apr 2026 Official leaderboard
30 GPT-5.6 Luna 61.3% Jul 2026 Official leaderboard
31 GPT-5.2 58.2% Dec 2025 Official leaderboard
32 DeepSeek V4.1 Flash 57.6% Sep 2026 Official leaderboard
33 Claude Opus 4.7 57.4% Apr 2026 Official leaderboard
34 Tencent Hy4 preview 56.9% Aug 2026 Official leaderboard
35 Claude Opus 4.8 55.8% May 2026 Official leaderboard
36 Kimi K3 55.7% Jul 2026 Official leaderboard
37 Claude Opus 4.6 54.4% Feb 2026 Official leaderboard
38 Claude Opus 4.5 54.2% Nov 2025 Official leaderboard
39 GPT-4o Mar 2025 54% Mar 2025 Official leaderboard
40 GLM-5.3-Flash 54% Aug 2026 Official leaderboard
41 Gemini 2.5 Flash Sep 2025 53.6% Sep 2025 Official leaderboard
42 GPT-4.1 53.2% Apr 2025 Official leaderboard
43 GLM-5.2 53% Jun 2026 Official leaderboard
44 Qwen 3.8 Max 52.8% Aug 2026 Official leaderboard
45 GLM-5.3 52.4% Aug 2026 Official leaderboard
46 GPT-5.4 52.1% Mar 2026 Official leaderboard
47 Claude Opus 4 51.6% May 2025 Official leaderboard
48 GPT-5.4 Mini 50.8% Mar 2026 Official leaderboard
49 Claude 3.7 Sonnet 50.7% Feb 2025 Official leaderboard
50 Qwen 3.6 Plus 50.7% Mar 2026 Official leaderboard
51 GLM-5 50.3% Feb 2026 Official leaderboard
52 Claude Sonnet 5.5 50.1% Sep 2026 Official leaderboard
53 Grok 3 49.9% Apr 2025 Official leaderboard
54 Gemini 2.0 Flash Feb 2025 48.6% Feb 2025 Official leaderboard
55 MiMo-V2.6 Pro 47.8% Sep 2026 Official leaderboard
56 Claude Sonnet 5 47.4% Jun 2026 Official leaderboard
57 MiniMax-M3 46.8% Jun 2026 Official leaderboard
58 Nemotron 3 Ultra 46.5% Jun 2026 Official leaderboard
59 GPT-4 Jun 2023 46.4% Jun 2023 Official leaderboard
60 Claude Sonnet 4.6 45.7% Feb 2026 Official leaderboard
61 Mistral Large 4 45.3% Oct 2026 Official leaderboard
62 Llama 4 Maverick 44.6% Apr 2025 Official leaderboard
63 Grok-3 mini 42.7% Apr 2025 Official leaderboard
64 MiMo-V2.6 Flash 42.3% Sep 2026 Official leaderboard
65 Claude Sonnet 4.5 42.1% Sep 2025 Official leaderboard
66 Kimi K2 Sep 2025 42% Sep 2025 Official leaderboard
67 Mistral Medium 3.5 42% Apr 2026 Official leaderboard
68 Qwen3.8 Flash 41.4% Aug 2026 Official leaderboard
69 Grok 4 Fast 40.9% Sep 2025 Official leaderboard
70 Claude Haiku 5.5 40.4% Oct 2026 Official leaderboard
71 Muse Glimmer 40% Aug 2026 Official leaderboard
72 Grok 4.20 39.5% Feb 2026 Official leaderboard
73 Claude Sonnet 4 39.4% May 2025 Official leaderboard
74 Claude 3.5 Haiku 39.2% Oct 2024 Official leaderboard
75 DeepSeek-V3 Mar 2025 39.2% Mar 2025 Official leaderboard
76 DeepSeek-V4-Flash 38.4% Apr 2026 Official leaderboard
77 DeepSeek-V3.2 37.8% Dec 2025 Official leaderboard
78 Step 3.7 Flash 37.3% May 2026 Official leaderboard
79 Mistral Small 4 37.2% Mar 2026 Official leaderboard
80 Llama 3.3 70B Instruct 35.7% Dec 2024 Official leaderboard
81 GPT-3.5 Turbo Jan 2024 35.3% Jan 2024 Official leaderboard
82 GPT-4o mini 35.1% Jul 2024 Official leaderboard
83 Qwen3.6 27B 35% Apr 2026 Official leaderboard
84 Claude Haiku 4.5 34.9% Oct 2025 Official leaderboard
85 Gemma 4 31B IT 34.7% Apr 2026 Official leaderboard
86 Gemma 3 27B 34.6% Mar 2025 Official leaderboard
87 Laguna S 2.1 34.5% Jul 2026 Official leaderboard
88 Claude 3 Haiku 32.6% Mar 2024 Official leaderboard
89 Qwen 3.8 27B 32.1% Aug 2026 Official leaderboard
90 Gemma 2 27B 32% Jun 2024 Official leaderboard
91 Qwen3.5-27B 31.9% Feb 2026 Official leaderboard
92 Step 3.5 Flash 25.7% Feb 2026 Official leaderboard
93 MiniMax-M2.7 23% Mar 2026 Official leaderboard
94 MiMo-V2-Pro 22.5% Mar 2026 Official leaderboard
95 GLM-4.7 21% Dec 2025 Official leaderboard
96 gpt-oss-120b 20.2% Aug 2025 Official leaderboard

2 models have held number one on the official leaderboard since its release in June 2026

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

Google DeepMind
Own leaderboard Passed 85% in Jun 2026, by Gemini 3.1 Pro

Source: Official EKI Eesti mäng leaderboard

Gemini 3.1 Pro on the official leaderboard passed 85% in February 2026

Models released before it came out already scored that high. 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 Best so far
0% 20% 40% 60% 80% 100% 2023 2024 2025 2026 2027 Near the ceiling (85%) Gemini 4 Argon Gemini 3.1 Pro Gemini 3 Flash Gemini 2.5 Pro GPT-4o Gemini 2.0 Flash GPT-4 0% 20% 40% 60% 80% 100% 2023 2024 2025 2026 2027 Near the ceiling (85%) Gemini 4 Argon Gemini 3 Flash

Source: Official EKI Eesti mäng 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 Multilingual

All 545 in Multilingual
  • MMMLU In use MMLU professionally translated into 14 languages. Multiple-choice accuracy. Claude Mythos Preview 92.7%
  • WMT24++ In use English-to-55-language translation of WMT24 test data. Scored with automatic metrics such as MetricX or xCOMET. K-EXAONE 90.5%
  • MMLU-ProX In use MMLU-Pro translated into up to 29 languages with expert review. Multiple-choice accuracy per language. Qwen3.7-Max 87%
  • CLIcK In use Models answer 1,995 multiple-choice questions on Korean language and culture. Questions come from official Korean exams and textbooks in eleven categories. Solar Open 2 90.7%
  • Global MMLU Lite 3.0 In use Models answer 400 human-translated MMLU questions per language in 23 languages. Version 3.0 adds Czech, Hungarian, Oriya, Slovak and Tajik and updates Italian. Accuracy. Claude Opus 5 95.7%
  • MMMB In use Models answer multiple-choice questions about images in six languages across 15 categories. The set has about 12,000 questions. LFM2.5-VL-3B 83%

Data audit

9 of 10 checks pass, 1 warning. Look first at the Task access check. We don’t know whether the tasks are public. Every benchmark

Catalog record 1 warning 1 pass
  • Warning Task access We don’t know whether the tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (101 entries).
  • Pass Models scored 96 models we track, plus 5 other entries on its leaderboard.
  • Pass Newest score The newest is from November 2026, for Claude Fable 5.2.
  • Pass Frontier models 7 of the top 10 have a score, led by Claude Fable 5.2 (#2).
  • Note Headroom Saturated. The best is 90%, past 85% since February 2026.
Leaderboard import 3 pass
  • Pass Importer We read the board on xn--mdupuu-pxaa.eki.ee: 96 rows for models we track and 5 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