Loading blueprint versions...
Please wait while we gather all the unique runs for this blueprint.
Please wait while we gather all the unique runs for this blueprint.
Please wait while we prepare the detailed comparison.
DTEF: Predict response distributions for Korea South. Source: Global Dialogues GD1 (https://github.com/collect-intel/global-dialogues)
Average key point coverage extent for each model across all prompts.
| Prompts vs. Models | Claude 3.7 Sonnet | Claude 3.5 Haiku | Claude Haiku 4.5 | Claude Sonnet 4 | Gemini 2.5 Flash | Gemma 3 12b It | Llama 3 70b Instruct | Llama 4 Maverick | Mistral Large 2411 | Mistral Medium 3 | Mistral Nemo | GPT 4.1 | GPT 4.1 Mini | GPT 4.1 Nano | GPT 4o | GPT 4o Mini | GPT OSS 120b | GPT OSS 20b | GLM 4.5 | Qwen3 30b A3B Instruct 2507 | Qwen3 32b | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Score | 11th 80.0% | 2nd 87.5% | 2nd 87.5% | 15th 77.3% | 20th 72.5% | 13th 77.5% | 5th 84.5% | 17th 77.0% | 19th 73.3% | 17th 77.0% | 1st 88.5% | 14th 77.5% | 4th 87.5% | 21st 69.0% | 15th 77.3% | 8th 81.8% | 12th 79.3% | 9th 81.0% | 7th 83.0% | 6th 84.0% | 10th 80.3% | |
| 88.9% | 82% | 97% | 99% | 75% | 81% | 77% | 93% | 81% | 87% | 92% | 96% | 78% | 96% | 84% | 98% | 96% | 90% | 90% | 94% | 97% | 84% | |
| 86.6% | 82% | 98% | 98% | 82% | 72% | 81% | 94% | 91% | 78% | 76% | 96% | 82% | 98% | 69% | 79% | 87% | 89% | 94% | 98% | 93% | 81% | |
| 67.8% | 74% | 74% | 74% | 74% | 66% | 74% | 72% | 62% | 60% | 63% | 77% | 68% | 74% | 51% | 58% | 67% | 63% | 63% | 66% | 72% | 72% | |
| 77.3% | 82% | 81% | 79% | 78% | 71% | 78% | 79% | 74% | 68% | 77% | 85% | 82% | 82% | 72% | 74% | 77% | 75% | 77% | 74% | 74% | 84% |