JEE/NEET LLM Benchmark - NEET 2026: leaderboard
The "NEET 2026" subtask of JEE/NEET LLM Benchmark. Vision-language benchmark over Indian JEE Advanced and NEET competitive-exam image questions, reporting exam scores for recent multimodal models across year and exam splits.
Metric: Exam Score (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 14 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash (Preview) | 99.31 | #82 |
| 2 | Gemini 3.1 Pro (Preview) | 98.61 | #58 |
| 3 | GPT-5.5 | 98.61 | #31 |
| 4 | Kimi K2.5 | 97.92 | #144 |
| 5 | Gemma 4 31B (IT) | 97.22 | #176 |
| 6 | Gemini 3.1 Flash Lite | 97.22 | #191 |
| 7 | Grok 4.3 | 96.67 | #188 |
| 8 | Qwen 3 VL 235B A22B (Thinking) | 95.14 | #238 (Qwen 3 VL 235B A22B) |
| 9 | Claude Sonnet 4.6 | 93.06 | #90 |
| 10 | GLM-4.6V | 89.58 | #316 |
| 11 | Claude Opus 4.7 | 84.03 | #49 |
| 12 | Claude Haiku 4.5 | 71.67 | #280 |
| 13 | GPT-5.4 | 69.03 | #79 |
| 14 | Mistral Medium 3.5 | 55.83 | #353 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=jee-neet-llm-benchmark-neet-2026 · How It Works · Data refreshed daily, snapshot 2026-10-20.