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 | 18th 75.5% | 5th 81.5% | 9th 80.3% | 14th 77.3% | 19th 74.5% | 16th 76.3% | 3rd 81.8% | 17th 75.8% | 21st 72.3% | 14th 77.3% | 4th 81.8% | 13th 77.8% | 1st 84.0% | 20th 73.5% | 5th 81.5% | 9th 80.3% | 8th 81.0% | 2nd 83.3% | 7th 81.0% | 9th 80.3% | 9th 80.3% | |
| 79.1% | 74% | 82% | 75% | 71% | 73% | 72% | 79% | 74% | 82% | 77% | 84% | 71% | 89% | 81% | 83% | 89% | 83% | 80% | 89% | 77% | 76% | |
| 72.9% | 60% | 76% | 78% | 63% | 69% | 70% | 76% | 74% | 64% | 64% | 72% | 69% | 78% | 58% | 82% | 72% | 83% | 97% | 75% | 75% | 75% | |
| 87.5% | 93% | 84% | 91% | 91% | 86% | 86% | 92% | 84% | 77% | 97% | 94% | 93% | 91% | 77% | 84% | 86% | 81% | 84% | 82% | 91% | 93% | |
| 76.1% | 75% | 84% | 77% | 84% | 70% | 77% | 80% | 71% | 66% | 71% | 77% | 78% | 78% | 78% | 77% | 74% | 77% | 72% | 78% | 78% | 77% |