Anthropic · JEE Advanced
Claude JEE Advanced score
Quick answer
Yes — on JEE Advanced. Claude Fable 5.1 scored 358/360 (99.4%) on the official 2026 papers, with one partial on Paper 2 Physics question 5. The same model scored a clean 360/360 on JEE Advanced 2025. Claude Sonnet 5 scored 341/360 in 2026.
These are Anthropic API models — Claude Fable 5.1 and Claude Sonnet 5 — sitting the printed English booklets. They did not see the official key. On Paper 2 Mathematics question 4, both Claude models wrote the value 1/3, so we restored plus 3 after the split image hid the A and B denominators.
Anthropic models on JEE Advanced 2026
| Model | Score | Maths | Physics | Chemistry |
|---|---|---|---|---|
| Claude Fable 5.1 | 358/360 | 120/120 | 118/120 | 120/120 |
| Claude Sonnet 5 | 341/360 | 120/120 | 118/120 | 103/120 |
Anthropic models on JEE Advanced 2025
| Model | Score | Percent |
|---|---|---|
| Claude Fable 5.1 | 360/360 | 100% |
Where Claude missed
Claude Fable 5.1 · 2026
- Paper 2 Physics Q5: official A, C, D, model A, D (2 / 4)
Claude Sonnet 5 · 2026
- Paper 1 Chemistry Q4: official C, model — (0 / 3)
- Paper 1 Chemistry Q8: official A, B, C, model A, C (2 / 4)
- Paper 1 Chemistry Q11: official 4, model 3 (0 / 4)
- Paper 1 Chemistry Q15: official C, model — (0 / 4)
- Paper 2 Chemistry Q2: official B, model D (−1 / 3)
- Paper 2 Physics Q5: official A, C, D, model A, D (2 / 4)
-
AI cracks JEE Advanced 2026
Four-model write-up and the question 4 rescore that set the ranking.
-
Compare with ChatGPT
GPT 6 Astra 356/360, GPT 5.5 351/360 on the same papers.
-
Can AI solve JEE?
The full 2025 and 2026 answer across providers.
-
Open the leaderboard
Heatmaps, paper splits, and every printed question.
Frequently asked questions
What is Claude’s JEE Advanced score?
On official JEE Advanced 2026, Claude Fable 5.1 358/360, Claude Sonnet 5 341/360. On JEE Advanced 2025, Claude Fable 5.1 scored 360/360.
Can Claude solve JEE?
Yes, on JEE Advanced. Claude Fable 5.1 scored 358 out of 360 (99.4%) in 2026, with one partial on Paper 2 Physics question 5. The same model scored a clean 360/360 on JEE Advanced 2025.
Is Claude better than ChatGPT on JEE Advanced?
On the 2026 official papers, Claude Fable 5.1 leads at 358/360, ahead of GPT 6 Astra at 356/360 and GPT 5.5 at 351/360. Claude Sonnet 5 is lower at 341/360. Rankings can move when we add models or years.
Did Claude see the official answers?
No. Each request included only the printed question image and its type. Official IIT keys were applied after the model replied.
Why was Claude credited on Paper 2 Mathematics question 4?
The split image hid the denominators under options A and B. Models that wrote the value 1/3 receive plus 3. Claude Fable 5.1 and Claude Sonnet 5 wrote 1/3. GPT 6 Astra and GPT 5.5 returned only letter A, so their minus 1 stands.