BRIDGE Medical Leaderboard - Few-Shot: leaderboard
Published Average Performance across 87 tasks for Few-Shot protocol; five-shot is the Few-Shot setting. No across-protocol aggregate.
Metric: Average Performance (%). Source: huggingface.co. Saturation forecast: Around October 2027. 112 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.8 27B (Non-reasoning) | 55.67 | #105 (Qwen 3.8 27B) |
| 2 | Gemini 1.5 Pro (002) | 55.51 | #349 |
| 3 | Gemma 4 31B (IT) | 54.88 | #176 |
| 4 | Gemini 2.5 Flash | 53.36 | #245 |
| 5 | Gemini 2.0 Flash (001) | 53.33 | #360 |
| 6 | Gemma 4 26B A4B (IT) | 53.22 | #297 |
| 7 | GPT-4o (2024-08-06) | 52.59 | #336 |
| 8 | DeepSeek R1 | 51.38 | #251 |
| 9 | Qwen 2.5 72B Instruct | 50.99 | #454 |
| 10 | Qwen 3 Next 80B A3B (Thinking) | 50.9 | #320 (Qwen 3 Next 80B A3B) |
| 11 | Mistral Large 2 (Nov) Instruct (2411) | 50.68 | #465 |
| 12 | Qwen 3 Next 80B A3B Instruct | 50.56 | #368 |
| 13 | Llama 3.1 70B Instruct | 50.52 | #561 |
| 14 | Qwen 3 30B A3B 2507 (Thinking) | 50.26 | #376 (Qwen 3 30B A3B 2507) |
| 15 | Qwen 2.5 32B Instruct | 49.54 | #498 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=bridge-medical-leaderboard-few-shot · How It Works · Data refreshed daily, snapshot 2026-10-20.