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 South Africa. Source: Global Dialogues GD5 (https://github.com/collect-intel/global-dialogues)
Average key point coverage extent for each model across all prompts.
| Prompts vs. Models | Claude Haiku 4.5 | Gemini 2.5 Flash | Gemma 3 12b It | GPT 4.1 Mini | GPT 4.1 Nano | GPT 4o Mini | |
|---|---|---|---|---|---|---|---|
| Score | 3rd 67.8% | 6th 52.8% | 5th 58.5% | 2nd 68.4% | 4th 63.2% | 1st 69.5% | |
| 74.5% | 82% | 67% | 76% | 84% | 63% | 75% | |
| 53.0% | 54% | 59% | 40% | 48% | 60% | 57% | |
| 77.7% | 78% | 80% | 67% | 81% | 77% | 83% | |
| 68.8% | 80% | 83% | 62% | 77% | 39% | 72% | |
| 50.3% | 52% | 66% | 37% | 54% | 47% | 46% | |
| 71.2% | 64% | 67% | 70% | 71% | 78% | 77% | |
| 74.3% | 84% | 0% | 76% | 95% | 91% | 100% | |
| 57.3% | 65% | 56% | 52% | 54% | 53% | 64% | |
| 58.5% | 57% | 50% | 59% | 62% | 60% | 63% | |
| 48.0% | 62% | 0% | 46% | 58% | 64% | 58% |