Benchmark Almanac

Icelandic WikiQA

Claude Fable 5.2 86.6% leads since November 2026
Not in model cards

Models answer 1,900 factual questions drawn from Icelandic Wikipedia. A language model grades each answer against a human-reviewed reference.

Passes test 2. Miðeind reports scores for 60 named models on its current leaderboard. The current judge is GPT-5.6 Luna. The archived board used GPT-4o.

Made by
Miðeind
Released
July 2024
Kind
Knowledge in many languages
Level
Everyday
Scored
Graded by a model
Answers
Open-ended
Tasks
All public
Measures
Capability
Language
Icelandic
In and out
Text → Text
Interaction
Single turn
Items from
Real work and use
Refresh
One fixed set
Run by
Anyone
Strongest model tested
Claude Fable 5.2 #2 on the capability index
Last model added
November 2026
Scores refreshed
November 20th, 2026
62
Entries on its own leaderboard, 59 models we track

Claude Fable 5.2 leads with 86.6% Claude Fable 5.2 leads with 86.6%

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 Claude Fable 5.2 high 86.6%
  2. 2 Claude Opus 5.5 high 74.3%
  3. 3 Claude Fable 5.1 high 74%
  4. 4 Gemini 4 Argon high 71.3%
  5. 5 Claude Opus 5 high 69.1%
  6. 6 Gemini 3.1 Pro high 68.9%
  7. 7 Gemini 4 Neon low 68.1%
  8. 8 Gemini 4 Carbon low 66.7%
  9. 9 GPT-6 Astra high 66.7%
  10. 10 Gemini 3.7 Flash high 66.6%
  11. 11 Gemini 3.8 Flash high 66.5%
  12. 12 Gemini 3.5 Flash high 66.4%
  13. 13 GPT-6.1 Sol high 65.9%
  14. 14 Gemini 4 Hydrogen low 65.5%
  15. 15 GPT-5.6 Sol high 64.5%
  16. 16 GPT-6.2 Sol high 64.5%
  17. 17 Claude Opus 4.8 high 62.7%
  18. 18 Claude Sonnet 5.5 high 59.8%
  19. 19 GPT-6 Sol high 59%
  20. 20 Gemini 2.5 Pro Jun 2025, high 56.2%
  21. 21 GPT-6.1 Terra high 53.9%
  22. 22 GPT-6 Luna high 50.7%
  23. 23 Muse Spark Open high 50.6%
  24. 24 Claude Sonnet 5 high 49.9%
  25. 25 Grok 4.7 low 49.1%
  26. 26 GPT-5.6 Luna high 47.8%
  27. 27 Inkling high 47.6%
  28. 28 GPT-5.6 Terra high 47.3%
  29. 29 DeepSeek-V4-Pro on 46%
  30. 30 Kimi K3 on 46%
  31. 31 Gemini 3.5 Flash-Lite high 45.7%
  32. 32 GPT-5.2 high 42.7%
  33. 33 DeepSeek V4.1 Flash on 41.6%
  34. 34 Inkling 2 Small minimal 39.4%
  35. 35 Claude Haiku 5.5 high 38.6%
  36. 36 GPT-4o Mar 2025, none 37.2%
  37. 37 Qwen3.8-2.4T-A95B on 35%
  38. 38 GLM-5.3-Flash on 35%
  39. 39 Mistral Large 3 none 30.5%
  40. 40 Mistral Large 4 high 28.6%
  41. 41 MiniMax-M3 on 27%
  42. 42 Inkling-Small high 26.5%
  43. 43 Gemini 2.5 Flash-Lite Sep 2025, high 25.2%
  44. 44 Llama 3.1 70B Instruct none 24%
  45. 45 Mistral Small 4 high 22.8%
  46. 46 Mistral Medium 3.5 high 22.4%
  47. 47 Gemma 4 31B IT on 19.3%
  48. 48 Gemma 4 26B A4B on 17.7%
  49. 49 Claude Haiku 4.5 off 16.9%
  50. 50 Apertus 1.5 70B on 14.1%
  51. 51 Bielik 11B v3.1 Instruct high 14%
  52. 52 GPT-3.5 Turbo Jan 2024, none 13.8%
  53. 53 Gemma 4 12B on 12.1%
  54. 54 MiniMax-M3.1-Flash high 11.8%
  55. 55 Qwen 3.8 27B xhigh 11.3%
  56. 56 Apertus 1.5 8B on 10.7%
  57. 57 GPT-4.1 nano none 8.58%
  58. 58 Llama 3.1 8B Instruct none 7.82%
  59. 59 Gemma 4 E4B on 7.47%
  60. 60 Apertus 8B Instruct none 6.39%
  61. 61 Gemma 4 E2B on 4.05%
  62. 62 Ripple-65B Preview on 3.37%

Source: Official Icelandic WikiQA leaderboard

  1. 1 Claude Fable 5.2 high 86.6%
  2. 2 Claude Opus 5.5 high 74.3%
  3. 3 Claude Fable 5.1 high 74%
  4. 4 Gemini 4 Argon high 71.3%
  5. 5 Claude Opus 5 high 69.1%
  6. 6 Gemini 3.1 Pro high 68.9%
  7. 7 Gemini 4 Neon low 68.1%
  8. 8 Gemini 4 Carbon low 66.7%
  9. 9 GPT-6 Astra high 66.7%
  10. 10 Gemini 3.7 Flash high 66.6%
  11. 11 Gemini 3.8 Flash high 66.5%
  12. 12 Gemini 3.5 Flash high 66.4%
  13. 13 GPT-6.1 Sol high 65.9%
  14. 14 Gemini 4 Hydrogen low 65.5%
  15. 15 GPT-5.6 Sol high 64.5%
  16. 16 GPT-6.2 Sol high 64.5%
  17. 17 Claude Opus 4.8 high 62.7%
  18. 18 Claude Sonnet 5.5 high 59.8%
  19. 19 GPT-6 Sol high 59%
  20. 20 Gemini 2.5 Pro Jun 2025, high 56.2%
  21. 21 GPT-6.1 Terra high 53.9%
  22. 22 GPT-6 Luna high 50.7%
  23. 23 Muse Spark Open high 50.6%
  24. 24 Claude Sonnet 5 high 49.9%
  25. 25 Grok 4.7 low 49.1%
  26. 26 GPT-5.6 Luna high 47.8%
  27. 27 Inkling high 47.6%
  28. 28 GPT-5.6 Terra high 47.3%
  29. 29 DeepSeek-V4-Pro on 46%
  30. 30 Kimi K3 on 46%
  31. 31 Gemini 3.5 Flash-Lite high 45.7%
  32. 32 GPT-5.2 high 42.7%
  33. 33 DeepSeek V4.1 Flash on 41.6%
  34. 34 Inkling 2 Small minimal 39.4%
  35. 35 Claude Haiku 5.5 high 38.6%
  36. 36 GPT-4o Mar 2025, none 37.2%
  37. 37 Qwen3.8-2.4T-A95B on 35%
  38. 38 GLM-5.3-Flash on 35%
  39. 39 Mistral Large 3 none 30.5%
  40. 40 Mistral Large 4 high 28.6%
  41. 41 MiniMax-M3 on 27%
  42. 42 Inkling-Small high 26.5%
  43. 43 Gemini 2.5 Flash-Lite Sep 2025, high 25.2%
  44. 44 Mistral Small 4 high 22.8%
  45. 45 Mistral Medium 3.5 high 22.4%
  46. 46 Gemma 4 31B IT on 19.3%
  47. 47 Gemma 4 26B A4B on 17.7%
  48. 48 Claude Haiku 4.5 off 16.9%
  49. 49 Apertus 1.5 70B on 14.1%
  50. 50 GPT-3.5 Turbo Jan 2024, none 13.8%
  51. 51 Gemma 4 12B on 12.1%
  52. 52 MiniMax-M3.1-Flash high 11.8%
  53. 53 Qwen 3.8 27B xhigh 11.3%
  54. 54 Apertus 1.5 8B on 10.7%
  55. 55 GPT-4.1 nano none 8.58%
  56. 56 Gemma 4 E4B on 7.47%
  57. 57 Apertus 8B Instruct none 6.39%
  58. 58 Gemma 4 E2B on 4.05%
  59. 59 Ripple-65B Preview on 3.37%

Source: Official Icelandic WikiQA leaderboard

#EntryScoreSettingReleasedRun onLinkSource
1 Claude Fable 5.2 86.6% high Nov 2026 Nov 2026 anthropic.com Official leaderboard
2 Claude Opus 5.5 74.3% high Sep 2026 Sep 2026 anthropic.com Official leaderboard
3 Claude Fable 5.1 74% high Sep 2026 Sep 2026 anthropic.com Official leaderboard
4 Gemini 4 Argon 71.3% high Sep 2026 Oct 2026 ai.google.dev Official leaderboard
5 Claude Opus 5 69.1% high Jul 2026 Sep 2026 anthropic.com Official leaderboard
6 Gemini 3.1 Pro 68.9% high Feb 2026 Sep 2026 ai.google.dev Official leaderboard
7 Gemini 4 Neon 68.1% low Nov 2026 Nov 2026 ai.google.dev Official leaderboard
8 Gemini 4 Carbon 66.7% low Oct 2026 Nov 2026 ai.google.dev Official leaderboard
9 GPT-6 Astra 66.7% high Sep 2026 Sep 2026 developers.openai.com Official leaderboard
10 Gemini 3.7 Flash 66.6% high Aug 2026 Sep 2026 ai.google.dev Official leaderboard
11 Gemini 3.8 Flash 66.5% high Sep 2026 Sep 2026 ai.google.dev Official leaderboard
12 Gemini 3.5 Flash 66.4% high May 2026 Sep 2026 ai.google.dev Official leaderboard
13 GPT-6.1 Sol 65.9% high Sep 2026 Sep 2026 developers.openai.com Official leaderboard
14 Gemini 4 Hydrogen 65.5% low Nov 2026 Nov 2026 ai.google.dev Official leaderboard
15 GPT-5.6 Sol 64.5% high Jul 2026 Sep 2026 developers.openai.com Official leaderboard
16 GPT-6.2 Sol 64.5% high Nov 2026 Nov 2026 developers.openai.com Official leaderboard
17 Claude Opus 4.8 62.7% high May 2026 Sep 2026 anthropic.com Official leaderboard
18 Claude Sonnet 5.5 59.8% high Sep 2026 Sep 2026 anthropic.com Official leaderboard
19 GPT-6 Sol 59% high Sep 2026 Sep 2026 developers.openai.com Official leaderboard
20 Gemini 2.5 Pro Jun 2025 56.2% high Jun 2025 Sep 2026 ai.google.dev Official leaderboard
21 GPT-6.1 Terra 53.9% high Oct 2026 Oct 2026 developers.openai.com Official leaderboard
22 GPT-6 Luna 50.7% high Sep 2026 Sep 2026 developers.openai.com Official leaderboard
23 Muse Spark Open 50.6% high Oct 2026 Oct 2026 huggingface.co Official leaderboard
24 Claude Sonnet 5 49.9% high Jun 2026 Sep 2026 anthropic.com Official leaderboard
25 Grok 4.7 49.1% low Sep 2026 Nov 2026 docs.x.ai Official leaderboard
26 GPT-5.6 Luna 47.8% high Jul 2026 Sep 2026 developers.openai.com Official leaderboard
27 Inkling 47.6% high Jul 2026 Sep 2026 huggingface.co Official leaderboard
28 GPT-5.6 Terra 47.3% high Jul 2026 Sep 2026 developers.openai.com Official leaderboard
29 DeepSeek-V4-Pro 46% on Apr 2026 Sep 2026 huggingface.co Official leaderboard
30 Kimi K3 46% on Jul 2026 Sep 2026 huggingface.co Official leaderboard
31 Gemini 3.5 Flash-Lite 45.7% high Jul 2026 Sep 2026 ai.google.dev Official leaderboard
32 GPT-5.2 42.7% high Dec 2025 Sep 2026 platform.openai.com Official leaderboard
33 DeepSeek V4.1 Flash 41.6% on Sep 2026 Oct 2026 huggingface.co Official leaderboard
34 Inkling 2 Small 39.4% minimal Nov 2026 Nov 2026 huggingface.co Official leaderboard
35 Claude Haiku 5.5 38.6% high Oct 2026 Oct 2026 anthropic.com Official leaderboard
36 GPT-4o Mar 2025 37.2% none Mar 2025 Sep 2026 platform.openai.com Official leaderboard
37 Qwen3.8-2.4T-A95B 35% on Aug 2026 Sep 2026 huggingface.co Official leaderboard
38 GLM-5.3-Flash 35% on Aug 2026 Sep 2026 huggingface.co Official leaderboard
39 Mistral Large 3 30.5% none Dec 2025 Sep 2026 mistral.ai Official leaderboard
40 Mistral Large 4 28.6% high Oct 2026 Oct 2026 mistral.ai Official leaderboard
41 MiniMax-M3 27% on Jun 2026 Sep 2026 huggingface.co Official leaderboard
42 Inkling-Small 26.5% high Jul 2026 Sep 2026 huggingface.co Official leaderboard
43 Gemini 2.5 Flash-Lite Sep 2025 25.2% high Sep 2025 Sep 2026 ai.google.dev Official leaderboard
44 Llama 3.1 70B Instruct 70B 24% none Oct 2026 huggingface.co Official leaderboard
45 Mistral Small 4 22.8% high Mar 2026 Sep 2026 mistral.ai Official leaderboard
46 Mistral Medium 3.5 22.4% high Apr 2026 Sep 2026 mistral.ai Official leaderboard
47 Gemma 4 31B IT 19.3% on Apr 2026 Sep 2026 deepmind.google Official leaderboard
48 Gemma 4 26B A4B 17.7% on Apr 2026 Oct 2026 huggingface.co Official leaderboard
49 Claude Haiku 4.5 16.9% off Oct 2025 Sep 2026 anthropic.com Official leaderboard
50 Apertus 1.5 70B 14.1% on Jul 2026 Sep 2026 huggingface.co Official leaderboard
51 Bielik 11B v3.1 Instruct 11.2B 14% high Nov 2026 huggingface.co Official leaderboard
52 GPT-3.5 Turbo Jan 2024 13.8% none Jan 2024 Sep 2026 platform.openai.com Official leaderboard
53 Gemma 4 12B 12.1% on Jun 2026 Nov 2026 huggingface.co Official leaderboard
54 MiniMax-M3.1-Flash 11.8% high Oct 2026 Oct 2026 huggingface.co Official leaderboard
55 Qwen 3.8 27B 11.3% xhigh Aug 2026 Oct 2026 huggingface.co Official leaderboard
56 Apertus 1.5 8B 10.7% on Jul 2026 Sep 2026 huggingface.co Official leaderboard
57 GPT-4.1 nano 8.58% none Apr 2025 Sep 2026 platform.openai.com Official leaderboard
58 Llama 3.1 8B Instruct 8B 7.82% none Oct 2026 huggingface.co Official leaderboard
59 Gemma 4 E4B 7.47% on Mar 2026 Oct 2026 huggingface.co Official leaderboard
60 Apertus 8B Instruct 6.39% none Sep 2025 Sep 2026 huggingface.co Official leaderboard
61 Gemma 4 E2B 4.05% on Mar 2026 Oct 2026 huggingface.co Official leaderboard
62 Ripple-65B Preview 3.37% on Nov 2026 Nov 2026 huggingface.co Official leaderboard
#ModelScoreSettingReleasedRun onLinkSource
1 Claude Fable 5.2 86.6% high Nov 2026 Nov 2026 anthropic.com Official leaderboard
2 Claude Opus 5.5 74.3% high Sep 2026 Sep 2026 anthropic.com Official leaderboard
3 Claude Fable 5.1 74% high Sep 2026 Sep 2026 anthropic.com Official leaderboard
4 Gemini 4 Argon 71.3% high Sep 2026 Oct 2026 ai.google.dev Official leaderboard
5 Claude Opus 5 69.1% high Jul 2026 Sep 2026 anthropic.com Official leaderboard
6 Gemini 3.1 Pro 68.9% high Feb 2026 Sep 2026 ai.google.dev Official leaderboard
7 Gemini 4 Neon 68.1% low Nov 2026 Nov 2026 ai.google.dev Official leaderboard
8 Gemini 4 Carbon 66.7% low Oct 2026 Nov 2026 ai.google.dev Official leaderboard
9 GPT-6 Astra 66.7% high Sep 2026 Sep 2026 developers.openai.com Official leaderboard
10 Gemini 3.7 Flash 66.6% high Aug 2026 Sep 2026 ai.google.dev Official leaderboard
11 Gemini 3.8 Flash 66.5% high Sep 2026 Sep 2026 ai.google.dev Official leaderboard
12 Gemini 3.5 Flash 66.4% high May 2026 Sep 2026 ai.google.dev Official leaderboard
13 GPT-6.1 Sol 65.9% high Sep 2026 Sep 2026 developers.openai.com Official leaderboard
14 Gemini 4 Hydrogen 65.5% low Nov 2026 Nov 2026 ai.google.dev Official leaderboard
15 GPT-5.6 Sol 64.5% high Jul 2026 Sep 2026 developers.openai.com Official leaderboard
16 GPT-6.2 Sol 64.5% high Nov 2026 Nov 2026 developers.openai.com Official leaderboard
17 Claude Opus 4.8 62.7% high May 2026 Sep 2026 anthropic.com Official leaderboard
18 Claude Sonnet 5.5 59.8% high Sep 2026 Sep 2026 anthropic.com Official leaderboard
19 GPT-6 Sol 59% high Sep 2026 Sep 2026 developers.openai.com Official leaderboard
20 Gemini 2.5 Pro Jun 2025 56.2% high Jun 2025 Sep 2026 ai.google.dev Official leaderboard
21 GPT-6.1 Terra 53.9% high Oct 2026 Oct 2026 developers.openai.com Official leaderboard
22 GPT-6 Luna 50.7% high Sep 2026 Sep 2026 developers.openai.com Official leaderboard
23 Muse Spark Open 50.6% high Oct 2026 Oct 2026 huggingface.co Official leaderboard
24 Claude Sonnet 5 49.9% high Jun 2026 Sep 2026 anthropic.com Official leaderboard
25 Grok 4.7 49.1% low Sep 2026 Nov 2026 docs.x.ai Official leaderboard
26 GPT-5.6 Luna 47.8% high Jul 2026 Sep 2026 developers.openai.com Official leaderboard
27 Inkling 47.6% high Jul 2026 Sep 2026 huggingface.co Official leaderboard
28 GPT-5.6 Terra 47.3% high Jul 2026 Sep 2026 developers.openai.com Official leaderboard
29 DeepSeek-V4-Pro 46% on Apr 2026 Sep 2026 huggingface.co Official leaderboard
30 Kimi K3 46% on Jul 2026 Sep 2026 huggingface.co Official leaderboard
31 Gemini 3.5 Flash-Lite 45.7% high Jul 2026 Sep 2026 ai.google.dev Official leaderboard
32 GPT-5.2 42.7% high Dec 2025 Sep 2026 platform.openai.com Official leaderboard
33 DeepSeek V4.1 Flash 41.6% on Sep 2026 Oct 2026 huggingface.co Official leaderboard
34 Inkling 2 Small 39.4% minimal Nov 2026 Nov 2026 huggingface.co Official leaderboard
35 Claude Haiku 5.5 38.6% high Oct 2026 Oct 2026 anthropic.com Official leaderboard
36 GPT-4o Mar 2025 37.2% none Mar 2025 Sep 2026 platform.openai.com Official leaderboard
37 Qwen3.8-2.4T-A95B 35% on Aug 2026 Sep 2026 huggingface.co Official leaderboard
38 GLM-5.3-Flash 35% on Aug 2026 Sep 2026 huggingface.co Official leaderboard
39 Mistral Large 3 30.5% none Dec 2025 Sep 2026 mistral.ai Official leaderboard
40 Mistral Large 4 28.6% high Oct 2026 Oct 2026 mistral.ai Official leaderboard
41 MiniMax-M3 27% on Jun 2026 Sep 2026 huggingface.co Official leaderboard
42 Inkling-Small 26.5% high Jul 2026 Sep 2026 huggingface.co Official leaderboard
43 Gemini 2.5 Flash-Lite Sep 2025 25.2% high Sep 2025 Sep 2026 ai.google.dev Official leaderboard
44 Mistral Small 4 22.8% high Mar 2026 Sep 2026 mistral.ai Official leaderboard
45 Mistral Medium 3.5 22.4% high Apr 2026 Sep 2026 mistral.ai Official leaderboard
46 Gemma 4 31B IT 19.3% on Apr 2026 Sep 2026 deepmind.google Official leaderboard
47 Gemma 4 26B A4B 17.7% on Apr 2026 Oct 2026 huggingface.co Official leaderboard
48 Claude Haiku 4.5 16.9% off Oct 2025 Sep 2026 anthropic.com Official leaderboard
49 Apertus 1.5 70B 14.1% on Jul 2026 Sep 2026 huggingface.co Official leaderboard
50 GPT-3.5 Turbo Jan 2024 13.8% none Jan 2024 Sep 2026 platform.openai.com Official leaderboard
51 Gemma 4 12B 12.1% on Jun 2026 Nov 2026 huggingface.co Official leaderboard
52 MiniMax-M3.1-Flash 11.8% high Oct 2026 Oct 2026 huggingface.co Official leaderboard
53 Qwen 3.8 27B 11.3% xhigh Aug 2026 Oct 2026 huggingface.co Official leaderboard
54 Apertus 1.5 8B 10.7% on Jul 2026 Sep 2026 huggingface.co Official leaderboard
55 GPT-4.1 nano 8.58% none Apr 2025 Sep 2026 platform.openai.com Official leaderboard
56 Gemma 4 E4B 7.47% on Mar 2026 Oct 2026 huggingface.co Official leaderboard
57 Apertus 8B Instruct 6.39% none Sep 2025 Sep 2026 huggingface.co Official leaderboard
58 Gemma 4 E2B 4.05% on Mar 2026 Oct 2026 huggingface.co Official leaderboard
59 Ripple-65B Preview 3.37% on Nov 2026 Nov 2026 huggingface.co Official leaderboard

More effort adds a median 1 points

Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.

Lowest effort to highest, light to dark
Claude Fable 5.2
low
86.6%
Gemini 4 Argon
off high
71.3%
Claude Opus 5
off high
69.1%
Gemini 3.1 Pro
low
68.9%
GPT-6 Astra
low
66.7%
Gemini 3.7 Flash
low
66.6%
Gemini 3.8 Flash
low
66.5%
Gemini 3.5 Flash
off high
66.4%
GPT-6.1 Sol
low
65.9%
Gemini 4 Hydrogen
low
65.5%
GPT-5.6 Sol
off high
64.5%
GPT-6.2 Sol
low high
64.5%

Source: Official Icelandic WikiQA leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Fable 5.2 high 86.6%
Claude Fable 5.2 low 84.4%
Claude Opus 5.5 high 74.3%
Claude Fable 5.1 high 74%
Gemini 4 Argon high 71.3%
Claude Opus 5 high 69.1%
Gemini 3.1 Pro high 68.9%
Gemini 4 Neon low 68.1%
Gemini 3.1 Pro low 67.9%
Gemini 4 Carbon low 66.7%
GPT-6 Astra high 66.7%
Gemini 3.7 Flash high 66.6%
GPT-6 Astra low 66.6%
Gemini 3.8 Flash high 66.5%
Gemini 3.5 Flash high 66.4%
GPT-6.1 Sol high 65.9%
Gemini 4 Argon off 65.6%
Gemini 4 Hydrogen low 65.5%
Gemini 3.7 Flash low 65.4%
Gemini 3.8 Flash low 65.4%
Gemini 4 Hydrogen high 65.2%
GPT-6.1 Sol low 64.9%
GPT-5.6 Sol high 64.5%
GPT-6.2 Sol high 64.5%
Claude Opus 5 off 63.2%
Claude Opus 4.8 high 62.7%
GPT-6.2 Sol low 60.4%
Claude Sonnet 5.5 high 59.8%
GPT-6 Sol high 59%
Claude Opus 4.8 off 57.9%
Gemini 3.5 Flash off 57.6%
Gemini 2.5 Pro Jun 2025 high 56.2%
GPT-5.6 Sol off 54.7%
GPT-6.1 Terra high 53.9%
Claude Sonnet 5.5 off 53.3%
GPT-6 Sol off 52.5%
GPT-6 Luna high 50.7%
Muse Spark Open high 50.6%
Gemini 2.5 Pro Jun 2025 minimal 50.1%
Claude Sonnet 5 high 49.9%
Claude Sonnet 5 off 49.3%
Grok 4.7 low 49.1%
GPT-5.6 Luna high 47.8%
Inkling high 47.6%
Grok 4.7 high 47.5%
GPT-5.6 Terra high 47.3%
GPT-6.1 Terra off 46.8%
DeepSeek-V4-Pro on 46%
Kimi K3 on 46%
Gemini 3.5 Flash-Lite high 45.7%
GPT-5.2 high 42.7%
DeepSeek-V4-Pro off 41.7%
DeepSeek V4.1 Flash on 41.6%
GPT-5.6 Terra off 41.5%
GPT-6 Luna off 40.5%
Inkling 2 Small minimal 39.4%
Gemini 3.5 Flash-Lite off 38.9%
Claude Haiku 5.5 high 38.6%
Inkling 2 Small high 38.6%
GPT-5.6 Luna off 37.6%
GPT-4o Mar 2025 none 37.2%
GPT-5.2 off 35.5%
Muse Spark Open off 35.3%
Qwen3.8-2.4T-A95B on 35%
GLM-5.3-Flash on 35%
Inkling off 34.7%
Kimi K3 off 31%
DeepSeek V4.1 Flash off 30.8%
Claude Haiku 5.5 off 30.8%
Mistral Large 3 none 30.5%
Mistral Large 4 high 28.6%
Mistral Large 4 low 28%
MiniMax-M3 on 27%
Inkling-Small high 26.5%
Gemini 2.5 Flash-Lite Sep 2025 high 25.2%
Mistral Small 4 high 22.8%
Mistral Medium 3.5 high 22.4%
Gemini 2.5 Flash-Lite Sep 2025 off 21.4%
Mistral Medium 3.5 off 19.7%
Mistral Small 4 off 19.3%
Gemma 4 31B IT on 19.3%
Gemma 4 26B A4B on 17.7%
Claude Haiku 4.5 off 16.9%
Claude Haiku 4.5 high 15.9%
Apertus 1.5 70B on 14.1%
GPT-3.5 Turbo Jan 2024 none 13.8%
Gemma 4 26B A4B off 12.6%
Gemma 4 12B on 12.1%
MiniMax-M3.1-Flash high 11.8%
Qwen 3.8 27B xhigh 11.3%
Apertus 1.5 8B on 10.7%
GPT-4.1 nano none 8.58%
Gemma 4 12B off 7.87%
Gemma 4 E4B on 7.47%
Qwen 3.8 27B off 7.18%
Apertus 8B Instruct none 6.39%
MiniMax-M3.1-Flash off 5.89%
Gemma 4 E4B off 4.97%
Gemma 4 E2B on 4.05%
Ripple-65B Preview on 3.37%
Gemma 4 E2B off 3.18%

More effort adds a median 1 points

Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.

Lowest effort to highest, light to dark
Claude Fable 5.2
low
86.6%
Gemini 4 Argon
off high
71.3%
Claude Opus 5
off high
69.1%
Gemini 3.1 Pro
low
68.9%
GPT-6 Astra
low
66.7%
Gemini 3.7 Flash
low
66.6%
Gemini 3.8 Flash
low
66.5%
Gemini 3.5 Flash
off high
66.4%
GPT-6.1 Sol
low
65.9%
Gemini 4 Hydrogen
low
65.5%
GPT-5.6 Sol
off high
64.5%
GPT-6.2 Sol
low high
64.5%

Source: Official Icelandic WikiQA leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Fable 5.2 high 86.6%
Claude Fable 5.2 low 84.4%
Claude Opus 5.5 high 74.3%
Claude Fable 5.1 high 74%
Gemini 4 Argon high 71.3%
Claude Opus 5 high 69.1%
Gemini 3.1 Pro high 68.9%
Gemini 4 Neon low 68.1%
Gemini 3.1 Pro low 67.9%
Gemini 4 Carbon low 66.7%
GPT-6 Astra high 66.7%
Gemini 3.7 Flash high 66.6%
GPT-6 Astra low 66.6%
Gemini 3.8 Flash high 66.5%
Gemini 3.5 Flash high 66.4%
GPT-6.1 Sol high 65.9%
Gemini 4 Argon off 65.6%
Gemini 4 Hydrogen low 65.5%
Gemini 3.7 Flash low 65.4%
Gemini 3.8 Flash low 65.4%
Gemini 4 Hydrogen high 65.2%
GPT-6.1 Sol low 64.9%
GPT-5.6 Sol high 64.5%
GPT-6.2 Sol high 64.5%
Claude Opus 5 off 63.2%
Claude Opus 4.8 high 62.7%
GPT-6.2 Sol low 60.4%
Claude Sonnet 5.5 high 59.8%
GPT-6 Sol high 59%
Claude Opus 4.8 off 57.9%
Gemini 3.5 Flash off 57.6%
Gemini 2.5 Pro Jun 2025 high 56.2%
GPT-5.6 Sol off 54.7%
GPT-6.1 Terra high 53.9%
Claude Sonnet 5.5 off 53.3%
GPT-6 Sol off 52.5%
GPT-6 Luna high 50.7%
Muse Spark Open high 50.6%
Gemini 2.5 Pro Jun 2025 minimal 50.1%
Claude Sonnet 5 high 49.9%
Claude Sonnet 5 off 49.3%
Grok 4.7 low 49.1%
GPT-5.6 Luna high 47.8%
Inkling high 47.6%
Grok 4.7 high 47.5%
GPT-5.6 Terra high 47.3%
GPT-6.1 Terra off 46.8%
DeepSeek-V4-Pro on 46%
Kimi K3 on 46%
Gemini 3.5 Flash-Lite high 45.7%
GPT-5.2 high 42.7%
DeepSeek-V4-Pro off 41.7%
DeepSeek V4.1 Flash on 41.6%
GPT-5.6 Terra off 41.5%
GPT-6 Luna off 40.5%
Inkling 2 Small minimal 39.4%
Gemini 3.5 Flash-Lite off 38.9%
Claude Haiku 5.5 high 38.6%
Inkling 2 Small high 38.6%
GPT-5.6 Luna off 37.6%
GPT-4o Mar 2025 none 37.2%
GPT-5.2 off 35.5%
Muse Spark Open off 35.3%
Qwen3.8-2.4T-A95B on 35%
GLM-5.3-Flash on 35%
Inkling off 34.7%
Kimi K3 off 31%
DeepSeek V4.1 Flash off 30.8%
Claude Haiku 5.5 off 30.8%
Mistral Large 3 none 30.5%
Mistral Large 4 high 28.6%
Mistral Large 4 low 28%
MiniMax-M3 on 27%
Inkling-Small high 26.5%
Gemini 2.5 Flash-Lite Sep 2025 high 25.2%
Mistral Small 4 high 22.8%
Mistral Medium 3.5 high 22.4%
Gemini 2.5 Flash-Lite Sep 2025 off 21.4%
Mistral Medium 3.5 off 19.7%
Mistral Small 4 off 19.3%
Gemma 4 31B IT on 19.3%
Gemma 4 26B A4B on 17.7%
Claude Haiku 4.5 off 16.9%
Claude Haiku 4.5 high 15.9%
Apertus 1.5 70B on 14.1%
GPT-3.5 Turbo Jan 2024 none 13.8%
Gemma 4 26B A4B off 12.6%
Gemma 4 12B on 12.1%
MiniMax-M3.1-Flash high 11.8%
Qwen 3.8 27B xhigh 11.3%
Apertus 1.5 8B on 10.7%
GPT-4.1 nano none 8.58%
Gemma 4 12B off 7.87%
Gemma 4 E4B on 7.47%
Qwen 3.8 27B off 7.18%
Apertus 8B Instruct none 6.39%
MiniMax-M3.1-Flash off 5.89%
Gemma 4 E4B off 4.97%
Gemma 4 E2B on 4.05%
Ripple-65B Preview on 3.37%
Gemma 4 E2B off 3.18%

8 models have held number one on the official leaderboard since its release in July 2024

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

OpenAI Google DeepMind Anthropic
Own leaderboard Passed 85% in Nov 2026, by Claude Fable 5.2

Source: Official Icelandic WikiQA leaderboard

Claude Fable 5.2 on the official leaderboard passed 85% in November 2026

That was 28 months after its release. By the same age, 38% of the benchmarks from 2024 had passed 85%. 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% 2024 2025 2026 2027 Near the ceiling (85%) Claude Fable 5.2 Claude Opus 5.5 Claude Opus 5 Gemini 2.5 Pro GPT-4o GPT-3.5 Turbo 0% 20% 40% 60% 80% 100% 2024 2025 2026 2027 Near the ceiling (85%) Claude Fable 5.2 Claude Opus 5.5

Source: Official Icelandic WikiQA 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

All 10 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (62 entries).
  • Pass Models scored 59 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 9 of the top 10 have a score, led by Claude Fable 5.2 (#2).
  • Note Headroom Saturated. The best is 86.6%, past 85% since November 2026.
Leaderboard import 3 pass
  • Pass Importer We read the board on mideind-icelandic-llm-leaderboard.static.hf.space: 101 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