Benchmark Almanac

Miðeind Icelandic Terminology

Claude Fable 5.2 78% leads since November 2026
Not in model cards

Models explain Icelandic knitting and crochet terms and name local species from English and Latin names. A model grades definitions. A program checks species names.

Passes test 2. Miðeind reports scores for 60 named models on its current leaderboard. The board combines its judged definitions and exact species names into one terminology score.

Made by
Miðeind
Released
September 2026
Kind
Knowledge in many languages
Level
University
Scored
Graded by a model
Answers
Short answer
Tasks
Kept private
Measures
Capability
Language
Icelandic, English, Latin
In and out
Text → Text
Interaction
Single turn
Items from
Written by experts
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
62
Entries on its own leaderboard, 61 models we track

Claude Fable 5.2 leads with 78% Claude Fable 5.2 leads with 78%

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 78%
  2. 2 Claude Opus 5.5 high 65.4%
  3. 3 GPT-6 Astra high 65.3%
  4. 4 GPT-6.1 Sol high 63.1%
  5. 5 Claude Fable 5.1 high 61.6%
  6. 6 GPT-6.2 Sol high 60.5%
  7. 7 Gemini 4 Argon high 60.2%
  8. 8 Gemini 4 Neon low 59.5%
  9. 9 Claude Opus 5 high 58.7%
  10. 10 GPT-5.6 Sol high 57.2%
  11. 11 Gemini 4 Carbon low 56.4%
  12. 12 Gemini 4 Hydrogen high 56.3%
  13. 13 Gemini 3.8 Flash high 56.1%
  14. 14 Gemini 3.1 Pro low 55.8%
  15. 15 GPT-6 Sol high 54.6%
  16. 16 Gemini 3.7 Flash high 53.7%
  17. 17 Claude Opus 4.8 high 53.4%
  18. 18 GPT-6.1 Terra high 53.2%
  19. 19 Claude Sonnet 5.5 high 50.8%
  20. 20 Gemini 3.5 Flash high 50.5%
  21. 21 GPT-6 Luna high 47%
  22. 22 Gemini 2.5 Pro Jun 2025, high 46.5%
  23. 23 DeepSeek-V4-Pro on 46.1%
  24. 24 Muse Spark Open high 44.2%
  25. 25 Grok 4.7 high 43.9%
  26. 26 GPT-5.6 Terra high 43.1%
  27. 27 GPT-5.6 Luna high 42.8%
  28. 28 Claude Sonnet 5 high 42.5%
  29. 29 Inkling 2 Small minimal 42.1%
  30. 30 DeepSeek V4.1 Flash on 40.2%
  31. 31 Inkling high 38.6%
  32. 32 Claude Haiku 5.5 high 38%
  33. 33 Gemini 3.5 Flash-Lite high 37.3%
  34. 34 GPT-5.2 high 37.1%
  35. 35 Kimi K3 on 36.3%
  36. 36 Qwen3.8-2.4T-A95B on 30.8%
  37. 37 MiniMax-M3 on 29.8%
  38. 38 GPT-4o Mar 2025, none 29.1%
  39. 39 Mistral Large 4 high 26.3%
  40. 40 GLM-5.3-Flash on 23.1%
  41. 41 Gemini 2.5 Flash-Lite Sep 2025, high 21.4%
  42. 42 Inkling-Small high 20.4%
  43. 43 Gemma 4 31B IT on 18.5%
  44. 44 Mistral Large 3 none 16.9%
  45. 45 Gemma 4 26B A4B on 11%
  46. 46 Claude Haiku 4.5 off 10.4%
  47. 47 Mistral Small 4 off 9.66%
  48. 48 Qwen 3.8 27B xhigh 9.60%
  49. 49 Bielik 11B v3.1 Instruct high 9.23%
  50. 50 Llama 3.1-70B none 7.66%
  51. 51 Mistral Medium 3.5 high 7.05%
  52. 52 Gemma 4 12B on 6.89%
  53. 53 GPT-4.1 nano none 6.65%
  54. 54 MiniMax-M3.1-Flash high 6.54%
  55. 55 Apertus 1.5 70B on 5.83%
  56. 56 GPT-3.5 Turbo Jan 2024, none 4.13%
  57. 57 Apertus 1.5 8B on 3.00%
  58. 58 Gemma 4 E4B on 2.80%
  59. 59 Gemma 4 E2B on 1.65%
  60. 60 Apertus 8B Instruct none 1.40%
  61. 61 Llama 3.1-8B none 0.85%
  62. 62 Ripple-65B Preview on 0.39%

Source: Official Miðeind Icelandic Terminology leaderboard

  1. 1 Claude Fable 5.2 high 78%
  2. 2 Claude Opus 5.5 high 65.4%
  3. 3 GPT-6 Astra high 65.3%
  4. 4 GPT-6.1 Sol high 63.1%
  5. 5 Claude Fable 5.1 high 61.6%
  6. 6 GPT-6.2 Sol high 60.5%
  7. 7 Gemini 4 Argon high 60.2%
  8. 8 Gemini 4 Neon low 59.5%
  9. 9 Claude Opus 5 high 58.7%
  10. 10 GPT-5.6 Sol high 57.2%
  11. 11 Gemini 4 Carbon low 56.4%
  12. 12 Gemini 4 Hydrogen high 56.3%
  13. 13 Gemini 3.8 Flash high 56.1%
  14. 14 Gemini 3.1 Pro low 55.8%
  15. 15 GPT-6 Sol high 54.6%
  16. 16 Gemini 3.7 Flash high 53.7%
  17. 17 Claude Opus 4.8 high 53.4%
  18. 18 GPT-6.1 Terra high 53.2%
  19. 19 Claude Sonnet 5.5 high 50.8%
  20. 20 Gemini 3.5 Flash high 50.5%
  21. 21 GPT-6 Luna high 47%
  22. 22 Gemini 2.5 Pro Jun 2025, high 46.5%
  23. 23 DeepSeek-V4-Pro on 46.1%
  24. 24 Muse Spark Open high 44.2%
  25. 25 Grok 4.7 high 43.9%
  26. 26 GPT-5.6 Terra high 43.1%
  27. 27 GPT-5.6 Luna high 42.8%
  28. 28 Claude Sonnet 5 high 42.5%
  29. 29 Inkling 2 Small minimal 42.1%
  30. 30 DeepSeek V4.1 Flash on 40.2%
  31. 31 Inkling high 38.6%
  32. 32 Claude Haiku 5.5 high 38%
  33. 33 Gemini 3.5 Flash-Lite high 37.3%
  34. 34 GPT-5.2 high 37.1%
  35. 35 Kimi K3 on 36.3%
  36. 36 Qwen3.8-2.4T-A95B on 30.8%
  37. 37 MiniMax-M3 on 29.8%
  38. 38 GPT-4o Mar 2025, none 29.1%
  39. 39 Mistral Large 4 high 26.3%
  40. 40 GLM-5.3-Flash on 23.1%
  41. 41 Gemini 2.5 Flash-Lite Sep 2025, high 21.4%
  42. 42 Inkling-Small high 20.4%
  43. 43 Gemma 4 31B IT on 18.5%
  44. 44 Mistral Large 3 none 16.9%
  45. 45 Gemma 4 26B A4B on 11%
  46. 46 Claude Haiku 4.5 off 10.4%
  47. 47 Mistral Small 4 off 9.66%
  48. 48 Qwen 3.8 27B xhigh 9.60%
  49. 49 Llama 3.1-70B none 7.66%
  50. 50 Mistral Medium 3.5 high 7.05%
  51. 51 Gemma 4 12B on 6.89%
  52. 52 GPT-4.1 nano none 6.65%
  53. 53 MiniMax-M3.1-Flash high 6.54%
  54. 54 Apertus 1.5 70B on 5.83%
  55. 55 GPT-3.5 Turbo Jan 2024, none 4.13%
  56. 56 Apertus 1.5 8B on 3.00%
  57. 57 Gemma 4 E4B on 2.80%
  58. 58 Gemma 4 E2B on 1.65%
  59. 59 Apertus 8B Instruct none 1.40%
  60. 60 Llama 3.1-8B none 0.85%
  61. 61 Ripple-65B Preview on 0.39%

Source: Official Miðeind Icelandic Terminology leaderboard

#EntryScoreSettingReleasedRun onLinkSource
1 Claude Fable 5.2 78% high Nov 2026 Nov 2026 anthropic.com Official leaderboard
2 Claude Opus 5.5 65.4% high Sep 2026 Sep 2026 anthropic.com Official leaderboard
3 GPT-6 Astra 65.3% high Sep 2026 developers.openai.com Official leaderboard
4 GPT-6.1 Sol 63.1% high Sep 2026 Sep 2026 developers.openai.com Official leaderboard
5 Claude Fable 5.1 61.6% high Sep 2026 anthropic.com Official leaderboard
6 GPT-6.2 Sol 60.5% high Nov 2026 Nov 2026 developers.openai.com Official leaderboard
7 Gemini 4 Argon 60.2% high Sep 2026 Oct 2026 ai.google.dev Official leaderboard
8 Gemini 4 Neon 59.5% low Nov 2026 Nov 2026 ai.google.dev Official leaderboard
9 Claude Opus 5 58.7% high Jul 2026 anthropic.com Official leaderboard
10 GPT-5.6 Sol 57.2% high Jul 2026 developers.openai.com Official leaderboard
11 Gemini 4 Carbon 56.4% low Oct 2026 Nov 2026 ai.google.dev Official leaderboard
12 Gemini 4 Hydrogen 56.3% high Nov 2026 Nov 2026 ai.google.dev Official leaderboard
13 Gemini 3.8 Flash 56.1% high Sep 2026 Sep 2026 ai.google.dev Official leaderboard
14 Gemini 3.1 Pro 55.8% low Feb 2026 Sep 2026 ai.google.dev Official leaderboard
15 GPT-6 Sol 54.6% high Sep 2026 Sep 2026 developers.openai.com Official leaderboard
16 Gemini 3.7 Flash 53.7% high Aug 2026 ai.google.dev Official leaderboard
17 Claude Opus 4.8 53.4% high May 2026 anthropic.com Official leaderboard
18 GPT-6.1 Terra 53.2% high Oct 2026 Oct 2026 developers.openai.com Official leaderboard
19 Claude Sonnet 5.5 50.8% high Sep 2026 Sep 2026 anthropic.com Official leaderboard
20 Gemini 3.5 Flash 50.5% high May 2026 ai.google.dev Official leaderboard
21 GPT-6 Luna 47% high Sep 2026 Sep 2026 developers.openai.com Official leaderboard
22 Gemini 2.5 Pro Jun 2025 46.5% high Jun 2025 Sep 2026 ai.google.dev Official leaderboard
23 DeepSeek-V4-Pro 46.1% on Apr 2026 Sep 2026 huggingface.co Official leaderboard
24 Muse Spark Open 44.2% high Oct 2026 Oct 2026 huggingface.co Official leaderboard
25 Grok 4.7 43.9% high Sep 2026 Nov 2026 docs.x.ai Official leaderboard
26 GPT-5.6 Terra 43.1% high Jul 2026 developers.openai.com Official leaderboard
27 GPT-5.6 Luna 42.8% high Jul 2026 developers.openai.com Official leaderboard
28 Claude Sonnet 5 42.5% high Jun 2026 anthropic.com Official leaderboard
29 Inkling 2 Small 42.1% minimal Nov 2026 Nov 2026 huggingface.co Official leaderboard
30 DeepSeek V4.1 Flash 40.2% on Sep 2026 Oct 2026 huggingface.co Official leaderboard
31 Inkling 38.6% high Jul 2026 Sep 2026 huggingface.co Official leaderboard
32 Claude Haiku 5.5 38% high Oct 2026 Oct 2026 anthropic.com Official leaderboard
33 Gemini 3.5 Flash-Lite 37.3% high Jul 2026 ai.google.dev Official leaderboard
34 GPT-5.2 37.1% high Dec 2025 platform.openai.com Official leaderboard
35 Kimi K3 36.3% on Jul 2026 huggingface.co Official leaderboard
36 Qwen3.8-2.4T-A95B 30.8% on Aug 2026 huggingface.co Official leaderboard
37 MiniMax-M3 29.8% on Jun 2026 huggingface.co Official leaderboard
38 GPT-4o Mar 2025 29.1% none Mar 2025 Sep 2026 platform.openai.com Official leaderboard
39 Mistral Large 4 26.3% high Oct 2026 Oct 2026 mistral.ai Official leaderboard
40 GLM-5.3-Flash 23.1% on Aug 2026 huggingface.co Official leaderboard
41 Gemini 2.5 Flash-Lite Sep 2025 21.4% high Sep 2025 ai.google.dev Official leaderboard
42 Inkling-Small 20.4% high Jul 2026 Sep 2026 huggingface.co Official leaderboard
43 Gemma 4 31B IT 18.5% on Apr 2026 Sep 2026 deepmind.google Official leaderboard
44 Mistral Large 3 16.9% none Dec 2025 Sep 2026 mistral.ai Official leaderboard
45 Gemma 4 26B A4B 11% on Apr 2026 Oct 2026 huggingface.co Official leaderboard
46 Claude Haiku 4.5 10.4% off Oct 2025 Sep 2026 anthropic.com Official leaderboard
47 Mistral Small 4 9.66% off Mar 2026 Sep 2026 mistral.ai Official leaderboard
48 Qwen 3.8 27B 9.60% xhigh Aug 2026 Oct 2026 huggingface.co Official leaderboard
49 Bielik 11B v3.1 Instruct 11.2B 9.23% high Nov 2026 huggingface.co Official leaderboard
50 Llama 3.1-70B 7.66% none Jul 2024 Oct 2026 huggingface.co Official leaderboard
51 Mistral Medium 3.5 7.05% high Apr 2026 Sep 2026 mistral.ai Official leaderboard
52 Gemma 4 12B 6.89% on Jun 2026 Nov 2026 huggingface.co Official leaderboard
53 GPT-4.1 nano 6.65% none Apr 2025 platform.openai.com Official leaderboard
54 MiniMax-M3.1-Flash 6.54% high Oct 2026 Oct 2026 huggingface.co Official leaderboard
55 Apertus 1.5 70B 5.83% on Jul 2026 Sep 2026 huggingface.co Official leaderboard
56 GPT-3.5 Turbo Jan 2024 4.13% none Jan 2024 Sep 2026 platform.openai.com Official leaderboard
57 Apertus 1.5 8B 3.00% on Jul 2026 Sep 2026 huggingface.co Official leaderboard
58 Gemma 4 E4B 2.80% on Mar 2026 Oct 2026 huggingface.co Official leaderboard
59 Gemma 4 E2B 1.65% on Mar 2026 Oct 2026 huggingface.co Official leaderboard
60 Apertus 8B Instruct 1.40% none Sep 2025 Sep 2026 huggingface.co Official leaderboard
61 Llama 3.1-8B 0.85% none Jul 2024 Oct 2026 huggingface.co Official leaderboard
62 Ripple-65B Preview 0.39% on Nov 2026 Nov 2026 huggingface.co Official leaderboard
#ModelScoreSettingReleasedRun onLinkSource
1 Claude Fable 5.2 78% high Nov 2026 Nov 2026 anthropic.com Official leaderboard
2 Claude Opus 5.5 65.4% high Sep 2026 Sep 2026 anthropic.com Official leaderboard
3 GPT-6 Astra 65.3% high Sep 2026 developers.openai.com Official leaderboard
4 GPT-6.1 Sol 63.1% high Sep 2026 Sep 2026 developers.openai.com Official leaderboard
5 Claude Fable 5.1 61.6% high Sep 2026 anthropic.com Official leaderboard
6 GPT-6.2 Sol 60.5% high Nov 2026 Nov 2026 developers.openai.com Official leaderboard
7 Gemini 4 Argon 60.2% high Sep 2026 Oct 2026 ai.google.dev Official leaderboard
8 Gemini 4 Neon 59.5% low Nov 2026 Nov 2026 ai.google.dev Official leaderboard
9 Claude Opus 5 58.7% high Jul 2026 anthropic.com Official leaderboard
10 GPT-5.6 Sol 57.2% high Jul 2026 developers.openai.com Official leaderboard
11 Gemini 4 Carbon 56.4% low Oct 2026 Nov 2026 ai.google.dev Official leaderboard
12 Gemini 4 Hydrogen 56.3% high Nov 2026 Nov 2026 ai.google.dev Official leaderboard
13 Gemini 3.8 Flash 56.1% high Sep 2026 Sep 2026 ai.google.dev Official leaderboard
14 Gemini 3.1 Pro 55.8% low Feb 2026 Sep 2026 ai.google.dev Official leaderboard
15 GPT-6 Sol 54.6% high Sep 2026 Sep 2026 developers.openai.com Official leaderboard
16 Gemini 3.7 Flash 53.7% high Aug 2026 ai.google.dev Official leaderboard
17 Claude Opus 4.8 53.4% high May 2026 anthropic.com Official leaderboard
18 GPT-6.1 Terra 53.2% high Oct 2026 Oct 2026 developers.openai.com Official leaderboard
19 Claude Sonnet 5.5 50.8% high Sep 2026 Sep 2026 anthropic.com Official leaderboard
20 Gemini 3.5 Flash 50.5% high May 2026 ai.google.dev Official leaderboard
21 GPT-6 Luna 47% high Sep 2026 Sep 2026 developers.openai.com Official leaderboard
22 Gemini 2.5 Pro Jun 2025 46.5% high Jun 2025 Sep 2026 ai.google.dev Official leaderboard
23 DeepSeek-V4-Pro 46.1% on Apr 2026 Sep 2026 huggingface.co Official leaderboard
24 Muse Spark Open 44.2% high Oct 2026 Oct 2026 huggingface.co Official leaderboard
25 Grok 4.7 43.9% high Sep 2026 Nov 2026 docs.x.ai Official leaderboard
26 GPT-5.6 Terra 43.1% high Jul 2026 developers.openai.com Official leaderboard
27 GPT-5.6 Luna 42.8% high Jul 2026 developers.openai.com Official leaderboard
28 Claude Sonnet 5 42.5% high Jun 2026 anthropic.com Official leaderboard
29 Inkling 2 Small 42.1% minimal Nov 2026 Nov 2026 huggingface.co Official leaderboard
30 DeepSeek V4.1 Flash 40.2% on Sep 2026 Oct 2026 huggingface.co Official leaderboard
31 Inkling 38.6% high Jul 2026 Sep 2026 huggingface.co Official leaderboard
32 Claude Haiku 5.5 38% high Oct 2026 Oct 2026 anthropic.com Official leaderboard
33 Gemini 3.5 Flash-Lite 37.3% high Jul 2026 ai.google.dev Official leaderboard
34 GPT-5.2 37.1% high Dec 2025 platform.openai.com Official leaderboard
35 Kimi K3 36.3% on Jul 2026 huggingface.co Official leaderboard
36 Qwen3.8-2.4T-A95B 30.8% on Aug 2026 huggingface.co Official leaderboard
37 MiniMax-M3 29.8% on Jun 2026 huggingface.co Official leaderboard
38 GPT-4o Mar 2025 29.1% none Mar 2025 Sep 2026 platform.openai.com Official leaderboard
39 Mistral Large 4 26.3% high Oct 2026 Oct 2026 mistral.ai Official leaderboard
40 GLM-5.3-Flash 23.1% on Aug 2026 huggingface.co Official leaderboard
41 Gemini 2.5 Flash-Lite Sep 2025 21.4% high Sep 2025 ai.google.dev Official leaderboard
42 Inkling-Small 20.4% high Jul 2026 Sep 2026 huggingface.co Official leaderboard
43 Gemma 4 31B IT 18.5% on Apr 2026 Sep 2026 deepmind.google Official leaderboard
44 Mistral Large 3 16.9% none Dec 2025 Sep 2026 mistral.ai Official leaderboard
45 Gemma 4 26B A4B 11% on Apr 2026 Oct 2026 huggingface.co Official leaderboard
46 Claude Haiku 4.5 10.4% off Oct 2025 Sep 2026 anthropic.com Official leaderboard
47 Mistral Small 4 9.66% off Mar 2026 Sep 2026 mistral.ai Official leaderboard
48 Qwen 3.8 27B 9.60% xhigh Aug 2026 Oct 2026 huggingface.co Official leaderboard
49 Llama 3.1-70B 7.66% none Jul 2024 Oct 2026 huggingface.co Official leaderboard
50 Mistral Medium 3.5 7.05% high Apr 2026 Sep 2026 mistral.ai Official leaderboard
51 Gemma 4 12B 6.89% on Jun 2026 Nov 2026 huggingface.co Official leaderboard
52 GPT-4.1 nano 6.65% none Apr 2025 platform.openai.com Official leaderboard
53 MiniMax-M3.1-Flash 6.54% high Oct 2026 Oct 2026 huggingface.co Official leaderboard
54 Apertus 1.5 70B 5.83% on Jul 2026 Sep 2026 huggingface.co Official leaderboard
55 GPT-3.5 Turbo Jan 2024 4.13% none Jan 2024 Sep 2026 platform.openai.com Official leaderboard
56 Apertus 1.5 8B 3.00% on Jul 2026 Sep 2026 huggingface.co Official leaderboard
57 Gemma 4 E4B 2.80% on Mar 2026 Oct 2026 huggingface.co Official leaderboard
58 Gemma 4 E2B 1.65% on Mar 2026 Oct 2026 huggingface.co Official leaderboard
59 Apertus 8B Instruct 1.40% none Sep 2025 Sep 2026 huggingface.co Official leaderboard
60 Llama 3.1-8B 0.85% none Jul 2024 Oct 2026 huggingface.co Official leaderboard
61 Ripple-65B Preview 0.39% on Nov 2026 Nov 2026 huggingface.co Official leaderboard

Scores at each effort level

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 high
78%
GPT-6 Astra
low
65.3%
GPT-6.1 Sol
low
63.1%
GPT-6.2 Sol
low high
60.5%
Gemini 4 Argon
off high
60.2%
Claude Opus 5
off high
58.7%
GPT-5.6 Sol
off high
57.2%
Gemini 4 Hydrogen
low
56.3%
Gemini 3.8 Flash
low high
56.1%
Gemini 3.1 Pro
low
55.8%
GPT-6 Sol
off high
54.6%
Gemini 3.7 Flash
low
53.7%

Source: Official Miðeind Icelandic Terminology leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Fable 5.2 high 78%
Claude Fable 5.2 low 74.9%
Claude Opus 5.5 high 65.4%
GPT-6 Astra high 65.3%
GPT-6 Astra low 64.4%
GPT-6.1 Sol high 63.1%
GPT-6.1 Sol low 62.4%
Claude Fable 5.1 high 61.6%
GPT-6.2 Sol high 60.5%
Gemini 4 Argon high 60.2%
Gemini 4 Neon low 59.5%
Claude Opus 5 high 58.7%
GPT-5.6 Sol high 57.2%
GPT-6.2 Sol low 56.5%
Gemini 4 Carbon low 56.4%
Gemini 4 Hydrogen high 56.3%
Gemini 3.8 Flash high 56.1%
Gemini 3.1 Pro low 55.8%
Gemini 3.1 Pro high 55.6%
Gemini 4 Argon off 55.5%
Gemini 4 Hydrogen low 55.3%
GPT-6 Sol high 54.6%
Gemini 3.7 Flash high 53.7%
Claude Opus 4.8 high 53.4%
GPT-6.1 Terra high 53.2%
Gemini 3.7 Flash low 52.8%
Gemini 3.8 Flash low 51.2%
Claude Sonnet 5.5 high 50.8%
Claude Opus 5 off 50.8%
Gemini 3.5 Flash high 50.5%
Claude Opus 4.8 off 49.4%
GPT-6 Sol off 48.6%
GPT-5.6 Sol off 48.5%
GPT-6 Luna high 47%
GPT-6.1 Terra off 46.8%
Gemini 2.5 Pro Jun 2025 high 46.5%
DeepSeek-V4-Pro on 46.1%
Muse Spark Open high 44.2%
Grok 4.7 high 43.9%
Grok 4.7 low 43.8%
DeepSeek-V4-Pro off 43.5%
GPT-5.6 Terra high 43.1%
GPT-5.6 Luna high 42.8%
Gemini 3.5 Flash off 42.6%
Claude Sonnet 5 high 42.5%
Claude Sonnet 5.5 off 42.5%
Gemini 2.5 Pro Jun 2025 minimal 42.5%
Inkling 2 Small minimal 42.1%
Claude Sonnet 5 off 41.4%
Inkling 2 Small high 41%
DeepSeek V4.1 Flash on 40.2%
Inkling high 38.6%
Claude Haiku 5.5 high 38%
Gemini 3.5 Flash-Lite high 37.3%
GPT-5.2 high 37.1%
GPT-6 Luna off 36.7%
Kimi K3 on 36.3%
GPT-5.6 Terra off 35.5%
GPT-5.6 Luna off 33.7%
GPT-5.2 off 33.5%
Muse Spark Open off 32.5%
DeepSeek V4.1 Flash off 30.8%
Qwen3.8-2.4T-A95B on 30.8%
MiniMax-M3 on 29.8%
Gemini 3.5 Flash-Lite off 29.8%
Claude Haiku 5.5 off 29.5%
Inkling off 29.3%
GPT-4o Mar 2025 none 29.1%
Mistral Large 4 high 26.3%
Kimi K3 off 25.8%
Mistral Large 4 low 24.3%
GLM-5.3-Flash on 23.1%
Gemini 2.5 Flash-Lite Sep 2025 high 21.4%
Inkling-Small high 20.4%
Gemma 4 31B IT on 18.5%
Mistral Large 3 none 16.9%
Gemini 2.5 Flash-Lite Sep 2025 off 15.4%
Gemma 4 26B A4B on 11%
Claude Haiku 4.5 off 10.4%
Claude Haiku 4.5 high 10.3%
Mistral Small 4 off 9.66%
Mistral Small 4 high 9.60%
Qwen 3.8 27B xhigh 9.60%
Llama 3.1-70B none 7.66%
Mistral Medium 3.5 high 7.05%
Gemma 4 12B on 6.89%
GPT-4.1 nano none 6.65%
MiniMax-M3.1-Flash high 6.54%
Mistral Medium 3.5 off 6.01%
Gemma 4 26B A4B off 5.88%
Apertus 1.5 70B on 5.83%
Qwen 3.8 27B off 4.30%
GPT-3.5 Turbo Jan 2024 none 4.13%
Gemma 4 12B off 3.50%
Apertus 1.5 8B on 3.00%
Gemma 4 E4B on 2.80%
MiniMax-M3.1-Flash off 2.44%
Gemma 4 E2B on 1.65%
Apertus 8B Instruct none 1.40%
Gemma 4 E4B off 1.30%
Llama 3.1-8B none 0.85%
Gemma 4 E2B off 0.70%
Ripple-65B Preview on 0.39%

Scores at each effort level

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 high
78%
GPT-6 Astra
low
65.3%
GPT-6.1 Sol
low
63.1%
GPT-6.2 Sol
low high
60.5%
Gemini 4 Argon
off high
60.2%
Claude Opus 5
off high
58.7%
GPT-5.6 Sol
off high
57.2%
Gemini 4 Hydrogen
low
56.3%
Gemini 3.8 Flash
low high
56.1%
Gemini 3.1 Pro
low
55.8%
GPT-6 Sol
off high
54.6%
Gemini 3.7 Flash
low
53.7%

Source: Official Miðeind Icelandic Terminology leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Fable 5.2 high 78%
Claude Fable 5.2 low 74.9%
Claude Opus 5.5 high 65.4%
GPT-6 Astra high 65.3%
GPT-6 Astra low 64.4%
GPT-6.1 Sol high 63.1%
GPT-6.1 Sol low 62.4%
Claude Fable 5.1 high 61.6%
GPT-6.2 Sol high 60.5%
Gemini 4 Argon high 60.2%
Gemini 4 Neon low 59.5%
Claude Opus 5 high 58.7%
GPT-5.6 Sol high 57.2%
GPT-6.2 Sol low 56.5%
Gemini 4 Carbon low 56.4%
Gemini 4 Hydrogen high 56.3%
Gemini 3.8 Flash high 56.1%
Gemini 3.1 Pro low 55.8%
Gemini 3.1 Pro high 55.6%
Gemini 4 Argon off 55.5%
Gemini 4 Hydrogen low 55.3%
GPT-6 Sol high 54.6%
Gemini 3.7 Flash high 53.7%
Claude Opus 4.8 high 53.4%
GPT-6.1 Terra high 53.2%
Gemini 3.7 Flash low 52.8%
Gemini 3.8 Flash low 51.2%
Claude Sonnet 5.5 high 50.8%
Claude Opus 5 off 50.8%
Gemini 3.5 Flash high 50.5%
Claude Opus 4.8 off 49.4%
GPT-6 Sol off 48.6%
GPT-5.6 Sol off 48.5%
GPT-6 Luna high 47%
GPT-6.1 Terra off 46.8%
Gemini 2.5 Pro Jun 2025 high 46.5%
DeepSeek-V4-Pro on 46.1%
Muse Spark Open high 44.2%
Grok 4.7 high 43.9%
Grok 4.7 low 43.8%
DeepSeek-V4-Pro off 43.5%
GPT-5.6 Terra high 43.1%
GPT-5.6 Luna high 42.8%
Gemini 3.5 Flash off 42.6%
Claude Sonnet 5 high 42.5%
Claude Sonnet 5.5 off 42.5%
Gemini 2.5 Pro Jun 2025 minimal 42.5%
Inkling 2 Small minimal 42.1%
Claude Sonnet 5 off 41.4%
Inkling 2 Small high 41%
DeepSeek V4.1 Flash on 40.2%
Inkling high 38.6%
Claude Haiku 5.5 high 38%
Gemini 3.5 Flash-Lite high 37.3%
GPT-5.2 high 37.1%
GPT-6 Luna off 36.7%
Kimi K3 on 36.3%
GPT-5.6 Terra off 35.5%
GPT-5.6 Luna off 33.7%
GPT-5.2 off 33.5%
Muse Spark Open off 32.5%
DeepSeek V4.1 Flash off 30.8%
Qwen3.8-2.4T-A95B on 30.8%
MiniMax-M3 on 29.8%
Gemini 3.5 Flash-Lite off 29.8%
Claude Haiku 5.5 off 29.5%
Inkling off 29.3%
GPT-4o Mar 2025 none 29.1%
Mistral Large 4 high 26.3%
Kimi K3 off 25.8%
Mistral Large 4 low 24.3%
GLM-5.3-Flash on 23.1%
Gemini 2.5 Flash-Lite Sep 2025 high 21.4%
Inkling-Small high 20.4%
Gemma 4 31B IT on 18.5%
Mistral Large 3 none 16.9%
Gemini 2.5 Flash-Lite Sep 2025 off 15.4%
Gemma 4 26B A4B on 11%
Claude Haiku 4.5 off 10.4%
Claude Haiku 4.5 high 10.3%
Mistral Small 4 off 9.66%
Mistral Small 4 high 9.60%
Qwen 3.8 27B xhigh 9.60%
Llama 3.1-70B none 7.66%
Mistral Medium 3.5 high 7.05%
Gemma 4 12B on 6.89%
GPT-4.1 nano none 6.65%
MiniMax-M3.1-Flash high 6.54%
Mistral Medium 3.5 off 6.01%
Gemma 4 26B A4B off 5.88%
Apertus 1.5 70B on 5.83%
Qwen 3.8 27B off 4.30%
GPT-3.5 Turbo Jan 2024 none 4.13%
Gemma 4 12B off 3.50%
Apertus 1.5 8B on 3.00%
Gemma 4 E4B on 2.80%
MiniMax-M3.1-Flash off 2.44%
Gemma 4 E2B on 1.65%
Apertus 8B Instruct none 1.40%
Gemma 4 E4B off 1.30%
Llama 3.1-8B none 0.85%
Gemma 4 E2B off 0.70%
Ripple-65B Preview on 0.39%

4 models have held number one on the official leaderboard since its release in September 2026

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

Anthropic OpenAI
Own leaderboard 4 leaders

Source: Official Miðeind Icelandic Terminology 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 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 Llama 3.1-70B 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 Miðeind Icelandic Terminology 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 11 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access The tasks are private.
  • Pass Confidence Checked against a primary source.
Scores 5 pass
  • Pass Independent runs Scores from the official leaderboard (62 entries).
  • Pass Models scored 61 models we track, plus 1 other entry 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).
  • Pass Headroom The best is 78%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on mideind-icelandic-llm-leaderboard.static.hf.space: 103 rows for models we track and 1 other entry 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