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 Philippines. 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 | 6th 76.8% | 15th 73.5% | 7th 76.8% | 9th 75.5% | 5th 77.3% | 18th 72.5% | 13th 74.0% | 4th 78.0% | 17th 73.0% | 11th 74.5% | 1st 79.0% | 3rd 78.3% | 10th 74.8% | 20th 70.0% | 7th 76.8% | 21st 69.8% | 19th 71.2% | 16th 73.3% | 12th 74.3% | 14th 73.5% | 2nd 78.5% | |
| 51.9% | 55% | 42% | 54% | 55% | 50% | 48% | 55% | 60% | 54% | 61% | 62% | 61% | 49% | 46% | 54% | 53% | 43% | 44% | 46% | 51% | 46% | |
| 82.1% | 84% | 85% | 84% | 78% | 97% | 81% | 75% | 85% | 74% | 69% | 81% | 83% | 84% | 74% | 95% | 78% | 83% | 83% | 78% | 78% | 95% | |
| 81.7% | 85% | 83% | 84% | 86% | 85% | 76% | 84% | 83% | 85% | 83% | 73% | 85% | 81% | 76% | 73% | 71% | 75% | 82% | 89% | 89% | 88% | |
| 83.6% | 83% | 84% | 85% | 83% | 77% | 85% | 82% | 84% | 79% | 85% | 100% | 84% | 85% | 84% | 85% | 77% | 84% | 84% | 84% | 76% | 85% |