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 Pakistan. 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 | 7th 83.0% | 3rd 84.8% | 8th 82.8% | 4th 84.3% | 19th 75.8% | 14th 78.5% | 1st 86.8% | 17th 77.3% | 16th 77.5% | 15th 78.3% | 2nd 85.0% | 9th 82.7% | 10th 81.5% | 20th 73.3% | 6th 83.2% | 11th 80.5% | 18th 76.5% | 13th 78.8% | 12th 79.8% | 5th 83.3% | 21st 73.0% | |
| 62.0% | 62% | 65% | 59% | 61% | 67% | 53% | 68% | 51% | 62% | 55% | 72% | 58% | 64% | 64% | 66% | 71% | 59% | 59% | 72% | 58% | 55% | |
| 88.8% | 93% | 96% | 86% | 93% | 81% | 87% | 96% | 90% | 92% | 89% | 91% | 97% | 86% | 76% | 97% | 87% | 86% | 83% | 93% | 94% | 72% | |
| 83.3% | 86% | 85% | 90% | 86% | 83% | 82% | 89% | 84% | 79% | 86% | 88% | 90% | 83% | 69% | 84% | 79% | 82% | 83% | 73% | 88% | 80% | |
| 87.1% | 91% | 93% | 96% | 97% | 72% | 92% | 94% | 84% | 77% | 83% | 89% | 86% | 93% | 84% | 86% | 85% | 79% | 90% | 81% | 93% | 85% |