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 Germany. 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 | 12th 77.5% | 1st 86.5% | 9th 79.8% | 14th 77.3% | 18th 73.5% | 18th 73.5% | 5th 81.0% | 17th 75.3% | 20th 72.5% | 15th 75.8% | 2nd 83.5% | 12th 77.5% | 3rd 83.2% | 21st 68.3% | 7th 80.3% | 5th 81.0% | 11th 79.0% | 16th 75.8% | 10th 79.0% | 4th 83.0% | 7th 80.3% | |
| 79.2% | 79% | 91% | 75% | 71% | 73% | 61% | 82% | 74% | 76% | 71% | 86% | 73% | 89% | 75% | 85% | 89% | 89% | 70% | 87% | 88% | 79% | |
| 70.6% | 60% | 78% | 76% | 63% | 64% | 70% | 76% | 74% | 64% | 64% | 74% | 66% | 78% | 53% | 82% | 75% | 72% | 74% | 69% | 75% | 75% | |
| 85.7% | 93% | 93% | 91% | 91% | 86% | 86% | 86% | 82% | 77% | 97% | 89% | 93% | 82% | 67% | 77% | 86% | 84% | 82% | 76% | 91% | 90% | |
| 77.6% | 78% | 84% | 77% | 84% | 71% | 77% | 80% | 71% | 73% | 71% | 85% | 78% | 84% | 78% | 77% | 74% | 71% | 77% | 84% | 78% | 77% |