MEGA-Bench Task - Exchange Rate Estimate Plot: leaderboard

Metric: Task Score (%). Source: huggingface.co. 44 models tracked.

Top models

#ModelScore
1Gemini 2.5 Pro (Preview 03-25)98.7
2Claude 3.5 Sonnet (20240620)98.4
3Qwen 2 VL 72B98.1
4Pixtral-12B98
5InternVL2.5-78B97.9
6GPT-4o97.6
7Claude 3.5 Sonnet (20241022)97.4
8Gemma 3 27B (IT)97.2
9Gemini 1.5 Pro (002)97.1
10Gemma 3 12B (IT)96.4
11Gemini 1.5 Flash (002)96.2
12GPT-4o Mini95.1
13Qwen 2 VL 7B94
14InternVL2-8B93.1
15Llama 4 Scout Base92.3

Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-exchange-rate-estimate-plot · How It Works · Data refreshed daily, snapshot 2026-09-05.