MarineEVT
Models answer 2,000 held-out questions about underwater events, including species interactions, movement, and timing. The authors report accuracy across five reasoning categories.
The authors publish scores for eight general video models without tools. They also compare three of those models with tools on the same questions.
- Released
- June 2026
- Kind
- Video understanding
- Level
- University
- Scored
- Checked automatically
- Answers
- Short answer
- Tasks
- All public
- Measures
- Capability
- Subject
- Biological sciences
- Language
- English
- Interaction
- Agentic
- Environment
- APIs and tools
- Items from
- Real work and use
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Agents, Vision, Biology and medicine
- Strongest model tested
-
Gemini 3 Flash #60 on the capability index
- Last model added
- December 2025
- Scores refreshed
- November 20th, 2026
Ours (EVT-R1) tops the leaderboard with 48.9% Qwen3-VL-8B-Instruct leads with 43.7% among models released by December 2025
Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart. 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 Qwen3-VL-8B-Instruct, at 43.7%. This view shows only models we track, each at its best.
The newest model with a score here came out in December 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
- 48.9 Ours (EVT-R1)
- 45.3 Qwen3-VL-8B (SFT)
- 43.7
Qwen3-VL-8B-Instruct
- 40.6
InternVL3-8B
- 40.4 GPT-5 mini
- 39.6 VideoITG
- 37.6
Gemini 3 Flash
- 35.9 CoF
- 35.6 Qwen3-VL-8B (GRPO†)
- 31.6 VideoLLAMA3-7B
- 30.6 Grok 4.1 Fast
- 30.4 MaxInfo
- 30.2 InternVL2-8B (PVC)
- 28.7 AKS
- 25.6 LLaVA-NeXT-Video-7B
- 23.8 LLaVA-1.5-7B (VisionZip)
- 21.9 Video-LLaVA-7B
- 20.4 Qwen2.5-VL-7B (VisionZip)
- 1 Ours (EVT-R1) Trained, Tools 48.9%
- 2 Qwen3-VL-8B (SFT) Trained, Tools 45.3%
- 3
Qwen3-VL-8B-Instruct No training, No tools 43.7%
- 4
InternVL3-8B No training, No tools 40.6%
- 5 GPT-5 mini No training, Tools 40.4%
- 6 VideoITG No training, No tools 39.6%
- 7
Gemini 3 Flash No training, No tools 37.6%
- 8 CoF No training, No tools 35.9%
- 9 Qwen3-VL-8B (GRPO†) Trained, Tools 35.6%
- 10 VideoLLAMA3-7B No training, No tools 31.6%
- 11 Grok 4.1 Fast reasoning, No training, No tools 30.6%
- 12 MaxInfo No training, No tools 30.4%
- 13 InternVL2-8B (PVC) No training, No tools 30.2%
- 14 AKS No training, No tools 28.7%
- 15 LLaVA-NeXT-Video-7B No training, No tools 25.6%
- 16 LLaVA-1.5-7B (VisionZip) No training, No tools 23.8%
- 17 Video-LLaVA-7B No training, No tools 21.9%
- 18 Qwen2.5-VL-7B (VisionZip) No training, No tools 20.4%
Source: Official MarineEVT leaderboard
- 1
Qwen3-VL-8B-Instruct No training, No tools 43.7%
- 2
InternVL3-8B No training, No tools 40.6%
- 3 GPT-5 mini No training, Tools 40.4%
- 4
Gemini 3 Flash No training, No tools 37.6%
- 5 VideoLLAMA3-7B No training, No tools 31.6%
- 6 Grok 4.1 Fast reasoning, No training, No tools 30.6%
- 7 Video-LLaVA-7B No training, No tools 21.9%
Source: Official MarineEVT leaderboard
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Ours (EVT-R1) open weights | 48.9% | Trained, Tools | Official leaderboard | |
| 2 | Qwen3-VL-8B (SFT) open weights | 45.3% | Trained, Tools | Official leaderboard | |
| 3 | | 43.7% | No training, No tools | Oct 2025 | Official leaderboard |
| 4 | | 40.6% | No training, No tools | Apr 2025 | Official leaderboard |
| 5 | GPT-5 mini | 40.4% | No training, Tools | Aug 2025 | Official leaderboard |
| 6 | VideoITG open weights | 39.6% | No training, No tools | Official leaderboard | |
| 7 | | 37.6% | No training, No tools | Dec 2025 | Official leaderboard |
| 8 | CoF open weights | 35.9% | No training, No tools | Official leaderboard | |
| 9 | Qwen3-VL-8B (GRPO†) open weights | 35.6% | Trained, Tools | Official leaderboard | |
| 10 | VideoLLAMA3-7B | 31.6% | No training, No tools | Jan 2024 | Official leaderboard |
| 11 | Grok 4.1 Fast | 30.6% | reasoning, No training, No tools | Nov 2025 | Official leaderboard |
| 12 | MaxInfo open weights | 30.4% | No training, No tools | Official leaderboard | |
| 13 | InternVL2-8B (PVC) open weights | 30.2% | No training, No tools | Official leaderboard | |
| 14 | AKS open weights | 28.7% | No training, No tools | Official leaderboard | |
| 15 | LLaVA-NeXT-Video-7B open weights | 25.6% | No training, No tools | Official leaderboard | |
| 16 | LLaVA-1.5-7B (VisionZip) open weights | 23.8% | No training, No tools | Official leaderboard | |
| 17 | Video-LLaVA-7B | 21.9% | No training, No tools | Nov 2023 | Official leaderboard |
| 18 | Qwen2.5-VL-7B (VisionZip) open weights | 20.4% | No training, No tools | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 43.7% | No training, No tools | Oct 2025 | Official leaderboard |
| 2 | | 40.6% | No training, No tools | Apr 2025 | Official leaderboard |
| 3 | GPT-5 mini | 40.4% | No training, Tools | Aug 2025 | Official leaderboard |
| 4 | | 37.6% | No training, No tools | Dec 2025 | Official leaderboard |
| 5 | VideoLLAMA3-7B | 31.6% | No training, No tools | Jan 2024 | Official leaderboard |
| 6 | Grok 4.1 Fast | 30.6% | reasoning, No training, No tools | Nov 2025 | Official leaderboard |
| 7 | Video-LLaVA-7B | 21.9% | No training, No tools | Nov 2023 | Official leaderboard |
Rankings by part
We read 5 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as Semantic Reasoning, Contextual Reasoning and Spatial Reasoning. Qwen3-VL-8B-Instruct leads 2 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.
Semantic Reasoning
7 models we track have a score here.
- 1 GPT-5 mini ↑2 58.4%
- 2
Qwen3-VL-8B-Instruct ↓1 58.2%
- 3
InternVL3-8B ↓1 53.4%
- 4
Gemini 3 Flash 48.2%
- 5 Grok 4.1 Fast ↑1 41.2%
- 6 VideoLLAMA3-7B ↓1 41%
- 7 Video-LLaVA-7B 24.4%
Contextual Reasoning
7 models we track have a score here.
- 1
Qwen3-VL-8B-Instruct 52.8%
- 2
InternVL3-8B 50.3%
- 3 VideoLLAMA3-7B ↑2 46.3%
- 4 GPT-5 mini ↓1 43.7%
- 5
Gemini 3 Flash ↓1 43.7%
- 6 Grok 4.1 Fast 37.3%
- 7 Video-LLaVA-7B 29%
Spatial Reasoning
7 models we track have a score here.
- 1
Gemini 3 Flash ↑3 27%
- 2 GPT-5 mini ↑1 22.8%
- 3
Qwen3-VL-8B-Instruct ↓2 22.4%
- 4 Grok 4.1 Fast ↑2 17.4%
- 5
InternVL3-8B ↓3 17.2%
- 6 Video-LLaVA-7B ↑1 8.40%
- 7 VideoLLAMA3-7B ↓2 5.40%
Temporal Reasoning
7 models we track have a score here.
- 1
Qwen3-VL-8B-Instruct 14%
- 2 GPT-5 mini ↑1 13%
- 3
InternVL3-8B ↓1 10.3%
- 4 Grok 4.1 Fast ↑2 7.75%
- 5
Gemini 3 Flash ↓1 7.75%
- 6 Video-LLaVA-7B ↑1 5.80%
- 7 VideoLLAMA3-7B ↓2 3.50%
Causal Reasoning
7 models we track have a score here.
- 1
InternVL3-8B ↑1 71.8%
- 2
Qwen3-VL-8B-Instruct ↓1 71%
- 3 GPT-5 mini 66.7%
- 4
Gemini 3 Flash 62%
- 5 VideoLLAMA3-7B 61.8%
- 6 Grok 4.1 Fast 50.7%
- 7 Video-LLaVA-7B 42%
Scores at each effort level
Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.
Source: Official MarineEVT leaderboard
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| No training, No tools | 43.7% | ||||
| No training, No tools | 40.6% | ||||
| GPT-5 mini | No training, Tools | 40.4% | |||
| GPT-5 mini | No training, No tools | 37.7% | |||
| No training, No tools | 37.6% | ||||
| No training, Tools | 36.3% | ||||
| VideoLLAMA3-7B | No training, No tools | 31.6% | |||
| Grok 4.1 Fast | reasoning, No training, No tools | 30.6% | |||
| Grok 4.1 Fast | reasoning, No training, Tools | 25.1% | |||
| Video-LLaVA-7B | No training, No tools | 21.9% |
Scores at each effort level
Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.
Source: Official MarineEVT leaderboard
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| No training, No tools | 43.7% | ||||
| No training, No tools | 40.6% | ||||
| GPT-5 mini | No training, Tools | 40.4% | |||
| GPT-5 mini | No training, No tools | 37.7% | |||
| No training, No tools | 37.6% | ||||
| No training, Tools | 36.3% | ||||
| VideoLLAMA3-7B | No training, No tools | 31.6% | |||
| Grok 4.1 Fast | reasoning, No training, No tools | 30.6% | |||
| Grok 4.1 Fast | reasoning, No training, Tools | 25.1% | |||
| Video-LLaVA-7B | No training, No tools | 21.9% |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official MarineEVT leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official MarineEVT 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 Audio and video
- Video-MMMU In use 300 expert-level lecture videos with questions testing perception, comprehension and adaptation of knowledge. Accuracy.
Qwen 3.8 Max 88.7%
- LVBench In use Hours-long video understanding with multiple-choice questions on reasoning, grounding and summarization.
Gemini 4 Argon 91.7%
- MMVU In use 3,000 expert-annotated questions on videos across 27 disciplines. Multiple-choice and open-ended accuracy.
Qwen 3.8 Max 82.4%
- MLVU In use Multi-task long video understanding (minutes to hours) with multiple-choice and generation tasks. Accuracy.
Qwen 3.8 Max 90.8%
- FLEURS In use Models transcribe read speech in 102 languages built from FLoRes-101 sentences. Model cards usually report word error rate, so lower scores are better.
Phi-4-multimodal 4.00%
- VoiceBench In use Spoken instructions with speaker, environment and content variation. Voice assistant responses scored by GPT judge or rules.
Qwen3.5-Omni-Plus 93.1%
Data audit
9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 11 months ago, in December 2025, for Gemini 3 Flash. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
- Pass Independent runs Scores from the official leaderboard (18 entries).
- Pass Models scored 7 models we track, plus 11 other entries on its leaderboard.
- Warning Newest score The newest came 11 months ago, in December 2025, for Gemini 3 Flash.
- Note Frontier models None of the top 10 has a score. The strongest tested is Gemini 3 Flash (#60).
- Pass Headroom The best is 43.7%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on marineevt.hkustvgd.com: 10 rows for models we track and 11 other entries 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.