InternVL2.5-8B
InternVL2.5-8B ranks first on 3 of the 43 benchmarks with five or more models scored.
Shanghai AI Lab Released December 5th, 2024 Family InternVL2.5 The lab's model card Compare with others 45
Benchmarks with a score
3
Ranked first, out of five or more models
7
In the top three
Shanghai AI Lab Dec 2024 InternVL2.5-8B
Open weights Small
No Epoch Capabilities Index score yet
- WorldSense (omnimodal) 39.1% #7 /12
- JointAVBench 51.7% #3 /6
- HumanEval-V 5.10% #20 /24
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 | 81.7% | #34 of 178 | Official leaderboard | |
| MM-IFEval | 46.2% | #4 of 9 | Official leaderboard | |
| Creation-MMBench | -25.42 | #12 of 19 | Official leaderboard |
Coding
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| HumanEval-V | 5.10% | #20 of 24 | Official leaderboard |
Agents and tools
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| UI-Vision Element Grounding | 2.09% | #3 of 5 | Official leaderboard |
Vision, audio and robotics
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| ChartQA Pro | 35.7% | #1 of 7 | Official leaderboard | |
| ViewSpatial-Bench | 43.2% | #1 of 7 | Official leaderboard | |
| MMBench-EN v1.1 | 83.2% | #29 of 178 | Official leaderboard | |
| SITE | 32.8% | #1 of 6 | Official leaderboard | |
| MMStar | 63.2% | #43 of 173 | Official leaderboard | |
| MM-Vet | 62.8% | #44 of 164 | Official leaderboard | |
| CoreCognition | 58.2% | #42 of 119 | Official leaderboard | |
| KRETA | 70.8% | #6 of 17 | Official leaderboard | |
| CharXiv (DQ) | 68.6% | #26 of 67 | Official leaderboard | |
| All-Angles Bench | 49.9% | #7 of 18 | Official leaderboard | |
| MMBench-Video | 1.68 | #11 of 28 | Official leaderboard | |
| MMSI-Bench | 28.7% | #11 of 28 | Official leaderboard | |
| ODV-Bench | 54.2% | #2 of 5 | Official leaderboard | |
| CharXiv | 32.9% | #37 of 77 | Official leaderboard | |
| JointAVBench | 51.7% | #3 of 6 | Official leaderboard | |
| CausalVQA | 47.5% | #2 of 4 | Official leaderboard | |
| MMMU | 56% | #71 of 136 | Official leaderboard | |
| SpaCE-10 | 41.8% | #19 of 36 | Official leaderboard | |
| LEGO-Puzzles | 35.2% | #9 of 16 | Official leaderboard | |
| WorldSense (omnimodal) | 39.1% | #7 of 12 | Official leaderboard | |
| VideoEval-Pro | 24.6% | #6 of 10 | Official leaderboard | |
| Do You See Me | 37% | #3 of 5 | Official leaderboard | |
| We-Math | 23.5% | #13 of 21 | Official leaderboard | |
| OST-Bench | 44.6% | #4 of 6 | Official leaderboard | |
| MMVU | 41.1% | #19 of 28 | Official leaderboard | |
| MMMU-Pro | 34.3% | #39 of 57 | Official leaderboard | |
| VRBench | 50.4% | #13 of 19 | Official leaderboard | |
| MME-CoT | 41.1% | #7 of 10 | Official leaderboard | |
| VSI-Bench-Debiased | 37.5% | #3 of 4 | Official leaderboard | |
| EmbodiedBench EB-Navigation | 21.3% | #14 of 17 | Official leaderboard | |
| MATH-Vision (MathVision) | 19.7% | #88 of 105 | Official leaderboard | |
| EmbodiedBench EB-Manipulation | 7.00% | #15 of 17 | Official leaderboard | |
| VCR-Bench | 23.9% | #8 of 9 | Official leaderboard | |
| EmoBench-M | 50.4% | #14 of 15 | Official leaderboard | |
| EmbodiedBench EB-ALFRED | 2.00% | #16 of 17 | Official leaderboard | |
| EmbodiedBench EB-Habitat | 11.3% | #17 of 17 | Official leaderboard | |
| MindCube | 18.7% | #7 of 7 | Official leaderboard | |
| Video-Holmes | 23.6% | #5 of 5 | Official leaderboard |
Safety and security
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Video-SafetyBench | 1.90 | #4 of 15 | Official leaderboard |
Other
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| OpenVLM Leaderboard | 68.1 | #43 of 173 | Official leaderboard |