DashBench
Models explain rare events in driving videos, including near misses and collisions. The authors compare 25 video models on 1,709 clips.
- Released
- November 2026
- Kind
- Video understanding
- Level
- Everyday
- Scored
- Graded by a model
- Answers
- Open-ended
- Tasks
- All public
- Measures
- Capability
- Language
- English
- Interaction
- Single turn
- Items from
- Real work and use
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Vision
- Strongest model tested
-
Gemini 3 Flash #60 on the capability index
- Last model added
- December 2025
- Scores refreshed
- November 20th, 2026
GPT-5 leads with 34.8% among models released by December 2025 GPT-5 leads with 34.8% 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. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. 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
- 34.8 GPT-5
- 34.4
Gemini 3 Flash
- 20.7
Qwen3-VL-235B-A22B
- 18.5 Qwen3-VL (8B)
- 17.2 Qwen3-VL (32B)
- 16.4
Ovis2.5-9B
- 16.1
InternVL3.5-241B-A28B
- 12.8
InternVL3.5-38B
- 12.6 Tarsier2 (7B)
- 12.3
MiniCPM-V 4.5
- 11.4
Keye-VL-8B-Preview
- 10.0 LLaVA-OV-RS (7B)
- 8.64
InternVL3.5-8B
- 7.76 InternVideo2.5 (8B)
- 7.74 IXC-2.5 (7B)
- 6.33 GLM-4.1V-9B-Thinking
- 5.91 CrashChat (7B)
- 5.44 VITA-1.5
- 4.38 LLaVA-OV-1.5 (8B)
- 2.76 DriveLMM-o1 (8B)
- 1.57 LLaVA-Next-Video (7B)
- 1.42 DriveMM (8B)
- 1.10 MiniInternVL2-BDD (4B)
- 0.87 MiniInternVL2-DLM (4B)
- 0.32 Dolphins (7B)
- 1 GPT-5 34.8%
- 2
Gemini 3 Flash 34.4%
- 3
Qwen3-VL-235B-A22B 20.7%
- 4 Qwen3-VL (8B) 18.5%
- 5 Qwen3-VL (32B) 17.2%
- 6
Ovis2.5-9B 16.4%
- 7
InternVL3.5-241B-A28B 16.1%
- 8
InternVL3.5-38B 12.8%
- 9 Tarsier2 (7B) 12.6%
- 10
MiniCPM-V 4.5 12.3%
- 11
Keye-VL-8B-Preview 11.4%
- 12 LLaVA-OV-RS (7B) 10%
- 13
InternVL3.5-8B 8.64%
- 14 InternVideo2.5 (8B) 7.76%
- 15 IXC-2.5 (7B) 7.74%
- 16 GLM-4.1V-9B-Thinking 6.33%
- 17 CrashChat (7B) 5.91%
- 18 VITA-1.5 5.44%
- 19 LLaVA-OV-1.5 (8B) 4.38%
- 20 DriveLMM-o1 (8B) 2.76%
- 21 LLaVA-Next-Video (7B) 1.57%
- 22 DriveMM (8B) 1.42%
- 23 MiniInternVL2-BDD (4B) 1.10%
- 24 MiniInternVL2-DLM (4B) 0.87%
- 25 Dolphins (7B) 0.32%
Source: Official DashBench leaderboard
- 1 GPT-5 34.8%
- 2
Gemini 3 Flash 34.4%
- 3
Qwen3-VL-235B-A22B 20.7%
- 4
Ovis2.5-9B 16.4%
- 5
InternVL3.5-241B-A28B 16.1%
- 6
InternVL3.5-38B 12.8%
- 7
MiniCPM-V 4.5 12.3%
- 8
Keye-VL-8B-Preview 11.4%
- 9
InternVL3.5-8B 8.64%
- 10 GLM-4.1V-9B-Thinking 6.33%
- 11 VITA-1.5 5.44%
Source: Official DashBench leaderboard
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-5 | 34.8% | Aug 2025 | Official leaderboard | |
| 2 | | 34.4% | Dec 2025 | Official leaderboard | |
| 3 | | 20.7% | Sep 2025 | Official leaderboard | |
| 4 | Qwen3-VL (8B) open weights | 18.5% | Official leaderboard | ||
| 5 | Qwen3-VL (32B) open weights | 17.2% | Official leaderboard | ||
| 6 | | 16.4% | Aug 2025 | Official leaderboard | |
| 7 | | 16.1% | Aug 2025 | Official leaderboard | |
| 8 | | 12.8% | Aug 2025 | Official leaderboard | |
| 9 | Tarsier2 (7B) open weights | 12.6% | Official leaderboard | ||
| 10 | | 12.3% | Aug 2025 | Official leaderboard | |
| 11 | | 11.4% | Jun 2025 | Official leaderboard | |
| 12 | LLaVA-OV-RS (7B) | 10% | Official leaderboard | ||
| 13 | | 8.64% | Aug 2025 | Official leaderboard | |
| 14 | InternVideo2.5 (8B) open weights | 7.76% | Official leaderboard | ||
| 15 | IXC-2.5 (7B) open weights | 7.74% | Official leaderboard | ||
| 16 | GLM-4.1V-9B-Thinking | 6.33% | Jul 2025 | Official leaderboard | |
| 17 | CrashChat (7B) | 5.91% | Official leaderboard | ||
| 18 | VITA-1.5 | 5.44% | Dec 2024 | Official leaderboard | |
| 19 | LLaVA-OV-1.5 (8B) open weights | 4.38% | Official leaderboard | ||
| 20 | DriveLMM-o1 (8B) | 2.76% | Official leaderboard | ||
| 21 | LLaVA-Next-Video (7B) open weights | 1.57% | Official leaderboard | ||
| 22 | DriveMM (8B) | 1.42% | Official leaderboard | ||
| 23 | MiniInternVL2-BDD (4B) | 1.10% | Official leaderboard | ||
| 24 | MiniInternVL2-DLM (4B) | 0.87% | Official leaderboard | ||
| 25 | Dolphins (7B) | 0.32% | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-5 | 34.8% | Aug 2025 | Official leaderboard | |
| 2 | | 34.4% | Dec 2025 | Official leaderboard | |
| 3 | | 20.7% | Sep 2025 | Official leaderboard | |
| 4 | | 16.4% | Aug 2025 | Official leaderboard | |
| 5 | | 16.1% | Aug 2025 | Official leaderboard | |
| 6 | | 12.8% | Aug 2025 | Official leaderboard | |
| 7 | | 12.3% | Aug 2025 | Official leaderboard | |
| 8 | | 11.4% | Jun 2025 | Official leaderboard | |
| 9 | | 8.64% | Aug 2025 | Official leaderboard | |
| 10 | GLM-4.1V-9B-Thinking | 6.33% | Jul 2025 | Official leaderboard | |
| 11 | VITA-1.5 | 5.44% | Dec 2024 | Official leaderboard |
Rankings by part
We read 6 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as Q1 (Unusual), Q2 (Dangerous) and Q3 (Describe). GPT-5 leads 5 of them, and Gemini 3 Flash leads the other one. Arrows show how far a model moved from its overall rank among the models we track.
Q1 (Unusual)
11 models we track have a score here.
- 1
Gemini 3 Flash ↑1 32.4%
- 2 GPT-5 ↓1 19.9%
- 3
Ovis2.5-9B ↑1 19.7%
- 4
InternVL3.5-241B-A28B ↑1 19.3%
- 5
Qwen3-VL-235B-A22B ↓2 18.7%
- 6
MiniCPM-V 4.5 ↑1 13.3%
- 7
InternVL3.5-38B ↓1 12.6%
- 8
Keye-VL-8B-Preview 12.4%
- 9 GLM-4.1V-9B-Thinking ↑1 7.85%
- 10 VITA-1.5 ↑1 7.31%
Q2 (Dangerous)
11 models we track have a score here.
- 1 GPT-5 36.9%
- 2
Gemini 3 Flash 35.3%
- 3
Qwen3-VL-235B-A22B 20.2%
- 4
Ovis2.5-9B 15.4%
- 5
InternVL3.5-38B ↑1 13.2%
- 6
InternVL3.5-241B-A28B ↓1 12.8%
- 7
MiniCPM-V 4.5 11.2%
- 8
Keye-VL-8B-Preview 11%
- 9
InternVL3.5-8B 7.67%
- 10 GLM-4.1V-9B-Thinking 5.50%
Q3 (Describe)
11 models we track have a score here.
- 1 GPT-5 31.2%
- 2
Gemini 3 Flash 29.8%
- 3
Qwen3-VL-235B-A22B 16.8%
- 4
InternVL3.5-241B-A28B ↑1 11.5%
- 5
InternVL3.5-38B ↑1 8.66%
- 6
Ovis2.5-9B ↓2 7.77%
- 7
InternVL3.5-8B ↑2 6.55%
- 8
MiniCPM-V 4.5 ↓1 6.38%
- 9
Keye-VL-8B-Preview ↓1 5.03%
- 10 VITA-1.5 ↑1 3.16%
Q4R (Reasoning)
11 models we track have a score here.
- 1 GPT-5 46.5%
- 2
Gemini 3 Flash 39.6%
- 3
Qwen3-VL-235B-A22B 25.9%
- 4
InternVL3.5-241B-A28B ↑1 19.1%
- 5
Ovis2.5-9B ↓1 18.7%
- 6
MiniCPM-V 4.5 ↑1 17.3%
- 7
InternVL3.5-38B ↓1 16.4%
- 8
Keye-VL-8B-Preview 13.1%
- 9
InternVL3.5-8B 12.2%
- 10 GLM-4.1V-9B-Thinking 6.32%
Q5 (Definition-guided)
11 models we track have a score here.
- 1 GPT-5 39.1%
- 2
Gemini 3 Flash 35%
- 3
Qwen3-VL-235B-A22B 21.7%
- 4
Ovis2.5-9B 20.6%
- 5
InternVL3.5-241B-A28B 17.6%
- 6
Keye-VL-8B-Preview ↑2 15.7%
- 7
MiniCPM-V 4.5 13.5%
- 8
InternVL3.5-38B ↓2 12.9%
- 9
InternVL3.5-8B 12.5%
- 10 GLM-4.1V-9B-Thinking 9.91%
Q4TF (Detection)
11 models we track have a score here.
- 1 GPT-5 97.8%
- 2
Gemini 3 Flash 94.8%
- 3
MiniCPM-V 4.5 ↑4 87.5%
- 4
Keye-VL-8B-Preview ↑4 80.1%
- 5
InternVL3.5-8B ↑4 78.2%
- 6
InternVL3.5-38B 77.7%
- 7
InternVL3.5-241B-A28B ↓2 76.4%
- 8
Qwen3-VL-235B-A22B ↓5 72%
- 9
Ovis2.5-9B ↓5 64.5%
- 10 VITA-1.5 ↑1 61.3%
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official DashBench 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 DashBench 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 (25 entries).
- Pass Models scored 11 models we track, plus 14 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 34.8%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on cornercase-dashbench.github.io: 11 rows for models we track and 14 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.