InternVL3-38B
InternVL3-38B places in the top three on ERQA.
13
Benchmarks with a score
0
Ranked first, out of five or more models
1
In the top three
Shanghai AI Lab Apr 2025 InternVL3-38B
Open weights Mid-size
No Epoch Capabilities Index score yet
- HiPhO 25.8% #31 /35
- A-Bench 77.1% #15 /116
- MMMU 70.1% #34 /136
Where it ranks
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
Knowledge and language
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| MMBench-CN v1.1 | 86.6% | #6 of 178 | Official leaderboard |
Math and science
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| HiPhO | 25.8% | #31 of 35 | Official leaderboard |
Vision, audio and robotics
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| MMStar | 72.6% | #5 of 173 | Official leaderboard | |
| MMBench | 87.7% | #5 of 155 | Official leaderboard | |
| MMBench-EN v1.1 | 87% | #6 of 178 | Official leaderboard | |
| CCBench (MMBench Chinese Culture) | 80.6% | #7 of 152 | Official leaderboard | |
| ERQA | 41.7% | #2 of 17 | Official leaderboard | |
| A-Bench | 77.1% | #15 of 116 | Official leaderboard | |
| CharXiv | 46.4% | #13 of 77 | Official leaderboard | |
| MMMU | 70.1% | #34 of 136 | Official leaderboard | |
| VisuLogic | 27.1% | #9 of 17 | Official leaderboard | |
| MATH-Vision (MathVision) | 34.5% | #68 of 105 | Official leaderboard |
Other
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| OpenVLM Leaderboard | 77.8 | #5 of 173 | Official leaderboard |