InternBootcamp
Models solve generated reasoning tasks with task-specific answer checkers. The hub port lets users select a task and sample examples with a fixed seed.
On August 26, 2026, the authors reported a best Bootcamp-Eval average of 61.1 for Qwen2.5-32B-Instruct with Bootcamp-SFT. Qwen3-235B-A22B scored 54.5.
- Made by
- InternLM
- Released
- May 2025
- Kind
- Language reasoning
- Level
- University
- Scored
- Checked automatically
- Answers
- Short answer
- Tasks
- All public
- Measures
- Capability
- Language
- English
- Interaction
- Single turn
- Items from
- Generated
- Refresh
- One fixed set
- Run by
- Anyone
- Strongest model tested
-
DeepSeek-R1 May 2025 #122 on the capability index
- Last model added
- May 2025
- Scores refreshed
- November 20th, 2026
Qwen2.5-32B-Instruct tops the leaderboard with 61.1 DeepSeek-R1-Distill-Qwen-32B leads with 55.7 among models released by May 2025
Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is DeepSeek-R1-Distill-Qwen-32B, at 55.7. This view shows only models we track, each at its best.
The newest model with a score here came out in May 2025, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 1 Qwen2.5-32B-Instruct Bootcamp-SFT 61.1
- 2
DeepSeek-R1-Distill-Qwen-32B Bootcamp-RL 55.7
- 3
Qwen3-235B-A22B 54.5
- 4
Qwen3-32B 52.1
- 5
QwQ-32B 51.4
- 6
DeepSeek-R1 May 2025 51.0
- 7
DeepSeek-V3 Mar 2025 46.5
Source: The InternBootcamp paper
- 1
DeepSeek-R1-Distill-Qwen-32B Bootcamp-RL 55.7
- 2
Qwen3-235B-A22B 54.5
- 3
Qwen3-32B 52.1
- 4
QwQ-32B 51.4
- 5
DeepSeek-R1 May 2025 51.0
- 6
DeepSeek-V3 Mar 2025 46.5
Source: The InternBootcamp paper
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Qwen2.5-32B-Instruct | 61.1 | Bootcamp-SFT | Official leaderboard | |
| 2 | | 55.7 | Bootcamp-RL | Jan 2025 | Official leaderboard |
| 3 | | 54.5 | Apr 2025 | Official leaderboard | |
| 4 | | 52.1 | Apr 2025 | Official leaderboard | |
| 5 | | 51.4 | Mar 2025 | Official leaderboard | |
| 6 | | 51.0 | May 2025 | Official leaderboard | |
| 7 | | 46.5 | Mar 2025 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 55.7 | Bootcamp-RL | Jan 2025 | Official leaderboard |
| 2 | | 54.5 | Apr 2025 | Official leaderboard | |
| 3 | | 52.1 | Apr 2025 | Official leaderboard | |
| 4 | | 51.4 | Mar 2025 | Official leaderboard | |
| 5 | | 51.0 | May 2025 | Official leaderboard | |
| 6 | | 46.5 | Mar 2025 | Official leaderboard |
Rankings by part
We read 8 parts from the benchmark’s paper, each ranking models on one part of the test, such as Graphical Puzzles, Algorithm and Logic. DeepSeek-R1-Distill-Qwen-32B leads 3 of them, and 3 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.
Graphical Puzzles
6 models we track have a score here.
- 1
Qwen3-235B-A22B ↑1 49.5
- 2
DeepSeek-R1 May 2025 ↑3 47.2
- 3
Qwen3-32B 45.9
- 4
QwQ-32B 45.2
- 5
DeepSeek-R1-Distill-Qwen-32B ↓4 41.8
- 6
DeepSeek-V3 Mar 2025 33.0
Algorithm
6 models we track have a score here.
- 1
DeepSeek-R1-Distill-Qwen-32B 61.3
- 2
DeepSeek-V3 Mar 2025 ↑4 57.4
- 3
Qwen3-235B-A22B ↓1 55.8
- 4
QwQ-32B 53.8
- 5
Qwen3-32B ↓2 53.7
- 6
DeepSeek-R1 May 2025 ↓1 46.6
Logic
6 models we track have a score here.
- 1
DeepSeek-R1-Distill-Qwen-32B 70.4
- 2
QwQ-32B ↑2 56.7
- 3
Qwen3-235B-A22B ↓1 54.5
- 4
Qwen3-32B ↓1 53.7
- 5
DeepSeek-R1 May 2025 50.9
- 6
DeepSeek-V3 Mar 2025 50.4
Cryptography
6 models we track have a score here.
- 1
DeepSeek-R1 May 2025 ↑4 61.1
- 2
Qwen3-235B-A22B 54.0
- 3
Qwen3-32B 45.8
- 4
QwQ-32B 37.6
- 5
DeepSeek-R1-Distill-Qwen-32B ↓4 36.7
- 6
DeepSeek-V3 Mar 2025 36.2
Math
6 models we track have a score here.
- 1
DeepSeek-R1-Distill-Qwen-32B 70.6
- 2
QwQ-32B ↑2 60.4
- 3
DeepSeek-R1 May 2025 ↑2 57.0
- 4
Qwen3-235B-A22B ↓2 56.9
- 5
Qwen3-32B ↓2 56.4
- 6
DeepSeek-V3 Mar 2025 50.7
Character Puzzles
6 models we track have a score here.
- 1
Qwen3-32B ↑2 81.5
- 2
Qwen3-235B-A22B 81.3
- 3
DeepSeek-R1-Distill-Qwen-32B ↓2 80.8
- 4
DeepSeek-R1 May 2025 ↑1 75.7
- 5
DeepSeek-V3 Mar 2025 ↑1 75.2
- 6
QwQ-32B ↓2 72.5
Natural Science
6 models we track have a score here.
- 1
Qwen3-32B ↑2 51.5
- 2
Qwen3-235B-A22B 49.1
- 3
DeepSeek-V3 Mar 2025 ↑3 46.8
- 4
DeepSeek-R1 May 2025 ↑1 45.8
- 5
DeepSeek-R1-Distill-Qwen-32B ↓4 45.1
- 6
QwQ-32B ↓2 42.3
Language Analysis
6 models we track have a score here.
- 1
DeepSeek-R1 May 2025 ↑4 67.7
- 2
Qwen3-235B-A22B 56.3
- 3
DeepSeek-R1-Distill-Qwen-32B ↓2 52.0
- 4
QwQ-32B 44.7
- 5
Qwen3-32B ↓2 39.7
- 6
DeepSeek-V3 Mar 2025 37.0
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: The InternBootcamp paper
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official InternBootcamp leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Knowledge
- MMLU-Pro In use Harder 10-option version of MMLU with reasoning-focused questions across 14 subjects.
Intern-S2-397B 89.8%
- MMLU In use 57-subject multiple-choice exam questions from elementary to professional level. Scored by accuracy.
Seed1.8 92.3%
- SuperGPQA In use 26,529 graduate-level multiple-choice questions across 285 disciplines. Accuracy.
Qwen 3.6 Max (Preview) 73.9%
- WideSearch In use 200 bilingual broad info-seeking tasks: agent collects many atomic facts into a table. Scored by success rate and row/item F1.
Tencent Hy4 preview 83.9%
- BBEH micro average In use Models solve harder replacements for each of the 23 BIG-Bench Hard tasks, 4,520 examples in total. The score is the share of all examples answered correctly.
Gemma 4 31B IT 74.4%
- HLE-Diamond In use Models answer 1,000 cleaned expert questions from Humanity's Last Exam, 500 on reasoning and 500 on knowledge, closed book. Graded on exact answers. Claude Fable 5.2 77.9%
Data audit
8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 18 months ago, in May 2025, for DeepSeek-R1. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
- Pass Independent runs Scores from the benchmark’s paper (7 entries).
- Pass Models scored 6 models we track, plus 1 other entry on its leaderboard.
- Warning Newest score The newest came 18 months ago, in May 2025, for DeepSeek-R1.
- Note Frontier models None of the top 10 has a score. The strongest tested is DeepSeek-R1 (#122).
- Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on arxiv.org: 7 rows for models we track and 1 other entry at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.