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 Bangladesh. 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 75.8% | 2nd 82.3% | 8th 78.3% | 18th 69.5% | 13th 74.3% | 17th 71.8% | 12th 75.0% | 3rd 79.8% | 13th 74.3% | 15th 73.8% | 9th 78.3% | 1st 83.5% | 3rd 79.8% | 21st 66.3% | 6th 79.0% | 10th 76.0% | 19th 68.8% | 20th 67.5% | 5th 79.3% | 16th 72.8% | 7th 78.8% | |
| 59.9% | 65% | 76% | 59% | 53% | 57% | 50% | 58% | 51% | 60% | 50% | 67% | 67% | 69% | 44% | 70% | 73% | 57% | 57% | 69% | 51% | 55% | |
| 73.1% | 82% | 82% | 80% | 81% | 61% | 78% | 67% | 81% | 61% | 65% | 77% | 76% | 82% | 69% | 69% | 61% | 69% | 75% | 72% | 73% | 75% | |
| 80.0% | 75% | 78% | 80% | 75% | 91% | 83% | 82% | 89% | 79% | 80% | 81% | 91% | 75% | 74% | 85% | 74% | 70% | 75% | 78% | 74% | 92% | |
| 88.7% | 81% | 93% | 94% | 69% | 88% | 76% | 93% | 98% | 97% | 100% | 88% | 100% | 93% | 78% | 92% | 96% | 79% | 63% | 98% | 93% | 93% |