Results by language
How each model scores for each language: the mean of its per-test scores, every test counted equally.
Items and answers come from the named sources and have not been reviewed by people, so a few gold answers may be wrong. How the items are built
Average across 9 models
#Averaged over 9 models, scores are highest for Welsh (85.1%) and lowest for Manx (65.2%).
Mean of each model's per-test scores for each language, averaged over models; highest first. Ranges per model are on each model page.
View as table
| Language | Average | Lowest model | Highest model |
|---|---|---|---|
| Welsh | 85.1% | Qwen3.8 Flash (69.9%) | Claude Fable 5.2 (91.5%) |
| Irish | 82.2% | Qwen3.8 Flash (69.1%) | Claude Fable 5.2 (89.8%) |
| Scottish Gaelic | 80.6% | Qwen3.8 Flash (71.2%) | Claude Fable 5.2 (86.6%) |
| Breton | 74.3% | Claude Haiku 5.5 (60.8%) | Claude Fable 5.2 (83.4%) |
| Cornish | 73.9% | Qwen3.8 Flash (49.6%) | Claude Fable 5.2 (86.3%) |
| Manx | 65.2% | Claude Haiku 5.5 (44.9%) | Claude Fable 5.2 (74.8%) |
Every model
#Darker cells are higher. Select a model for its ranges and item counts.
| Model | Irish | Welsh | Scottish Gaelic | Manx | Breton | Cornish |
|---|---|---|---|---|---|---|
| Claude Fable 5.2 | 89.8% | 91.5% | 86.6% | 74.8% | 83.4% | 86.3% |
| Gemini 4 Hydrogen | 87.3% | 89.4% | 86.1% | 74.3% | 80.9% | 83.1% |
| GPT-6 Astra | 86.6% | 89.2% | 83.8% | 72.3% | 82.4% | 82.9% |
| Gemini 4 Argon | 87.3% | 88.2% | 82.5% | 71.6% | 78.9% | 78.2% |
| DeepSeek V4.1 Flash 1017 | 81.3% | 86.7% | 81.8% | 68.6% | 77.2% | 78.9% |
| Kimi K3.1 Pro | 83.5% | 85.0% | 82.1% | 68.6% | 71.2% | 75.3% |
| GPT-6 Luna | 81.1% | 83.5% | 78.4% | 66.9% | 72.6% | 70.0% |
| Claude Haiku 5.5 | 73.6% | 82.7% | 73.2% | 44.9% | 60.8% | 60.5% |
| Qwen3.8 Flash | 69.1% | 69.9% | 71.2% | 45.0% | 61.5% | 49.6% |
Test by test
#Lesson exercises
| Model | Irish | Welsh | Scottish Gaelic | Manx | Breton | Cornish |
|---|---|---|---|---|---|---|
| Claude Fable 5.2 | 97.2% | 98.9% | 100.0% | 96.7% | 98.3% | 100.0% |
| Gemini 4 Hydrogen | 93.3% | 100.0% | 100.0% | 97.8% | 95.6% | 97.2% |
| GPT-6 Astra | 93.3% | 96.7% | 98.3% | 95.6% | 95.6% | 100.0% |
| Gemini 4 Argon | 95.0% | 99.4% | 98.3% | 99.4% | 98.3% | 98.9% |
| DeepSeek V4.1 Flash 1017 | 92.2% | 96.1% | 98.3% | 93.9% | 96.1% | 96.7% |
| Kimi K3.1 Pro | 93.3% | 95.6% | 97.2% | 94.4% | 96.7% | 97.8% |
| GPT-6 Luna | 90.0% | 94.4% | 96.1% | 95.6% | 94.4% | 91.1% |
| Claude Haiku 5.5 | 87.2% | 91.7% | 90.6% | 79.4% | 88.3% | 87.8% |
| Qwen3.8 Flash | 88.9% | 90.6% | 94.4% | 87.8% | 93.3% | 88.3% |
Word forms
| Model | Irish | Welsh | Scottish Gaelic | Manx | Breton | Cornish |
|---|---|---|---|---|---|---|
| Claude Fable 5.2 | 96.7% | 97.8% | 96.7% | 86.1% | 97.8% | 89.4% |
| Gemini 4 Hydrogen | 92.2% | 95.0% | 91.7% | 88.3% | 95.0% | 84.4% |
| GPT-6 Astra | 91.7% | 100.0% | 91.7% | 87.2% | 97.8% | 86.7% |
| Gemini 4 Argon | 88.3% | 91.1% | 87.8% | 82.8% | 87.2% | 75.0% |
| DeepSeek V4.1 Flash 1017 | 87.2% | 96.1% | 91.1% | 80.0% | 85.6% | 82.2% |
| Kimi K3.1 Pro | 90.6% | 92.2% | 87.8% | 80.6% | 67.2% | 74.4% |
| GPT-6 Luna | 87.2% | 90.0% | 86.7% | 78.9% | 72.2% | 67.8% |
| Claude Haiku 5.5 | 75.6% | 90.0% | 73.3% | 42.2% | 53.9% | 56.7% |
| Qwen3.8 Flash | 62.8% | 53.9% | 65.6% | 46.1% | 44.4% | 30.6% |
Translation
| Model | Irish | Welsh | Scottish Gaelic | Manx | Breton | Cornish |
|---|---|---|---|---|---|---|
| Claude Fable 5.2 | 82.0% | 83.8% | 77.3% | 52.5% | 71.9% | 83.0% |
| Gemini 4 Hydrogen | 81.4% | 79.1% | 76.0% | 49.7% | 69.3% | 79.1% |
| GPT-6 Astra | 80.8% | 80.8% | 77.0% | 49.6% | 71.6% | 78.9% |
| Gemini 4 Argon | 81.5% | 79.8% | 76.6% | 48.9% | 69.5% | 75.3% |
| DeepSeek V4.1 Flash 1017 | 76.9% | 76.3% | 76.4% | 46.2% | 65.6% | 71.2% |
| Kimi K3.1 Pro | 78.0% | 78.9% | 77.5% | 46.2% | 64.3% | 71.6% |
| GPT-6 Luna | 76.9% | 77.9% | 75.8% | 43.1% | 66.0% | 68.4% |
| Claude Haiku 5.5 | 69.7% | 77.1% | 70.5% | 26.5% | 52.9% | 50.3% |
| Qwen3.8 Flash | 64.9% | 70.7% | 74.9% | 24.5% | 58.3% | 48.3% |
Round trip
| Model | Irish | Welsh | Scottish Gaelic | Manx | Breton | Cornish |
|---|---|---|---|---|---|---|
| Claude Fable 5.2 | 82.0% | 83.6% | 77.8% | 64.1% | 65.7% | 72.9% |
| Gemini 4 Hydrogen | 82.1% | 83.6% | 74.9% | 61.5% | 63.9% | 71.8% |
| GPT-6 Astra | 77.7% | 79.6% | 70.8% | 56.8% | 64.5% | 66.1% |
| Gemini 4 Argon | 82.8% | 82.4% | 74.9% | 55.3% | 60.7% | 63.8% |
| DeepSeek V4.1 Flash 1017 | 74.1% | 81.0% | 74.8% | 54.3% | 61.4% | 65.3% |
| Kimi K3.1 Pro | 78.6% | 76.9% | 69.3% | 53.2% | 56.6% | 57.5% |
| GPT-6 Luna | 78.2% | 79.5% | 68.3% | 50.2% | 57.7% | 52.9% |
| Claude Haiku 5.5 | 68.9% | 83.0% | 69.5% | 31.6% | 47.9% | 47.2% |
| Qwen3.8 Flash | 65.5% | 75.3% | 69.2% | 21.7% | 50.2% | 31.2% |
Tutor accuracy
| Model | Irish | Welsh | Scottish Gaelic |
|---|---|---|---|
| Claude Fable 5.2 | 91.1% | 93.3% | 81.1% |
| Gemini 4 Hydrogen | 87.2% | 89.4% | 87.8% |
| GPT-6 Astra | 89.4% | 88.9% | 81.1% |
| Gemini 4 Argon | 88.9% | 88.3% | 75.0% |
| DeepSeek V4.1 Flash 1017 | 76.1% | 83.9% | 68.3% |
| Kimi K3.1 Pro | 77.2% | 81.7% | 78.9% |
| GPT-6 Luna | 73.3% | 75.6% | 65.0% |
| Claude Haiku 5.5 | 66.7% | 71.7% | 62.2% |
| Qwen3.8 Flash | 63.3% | 58.9% | 51.7% |