BabyVision
We show this table for reference; we do not rank on it.
A multimodal benchmark for fine-grained visual perception and grounded reasoning tasks.
Benchmark score on BabyVision — November 13, 2026
We compile the BabyVision rows from provider self-reports. Qwen3.8 Max leads the table at 82.0%, followed by Muse Spark Open (80.2%) and Muse Spark 1.1 (76.3%). We do not use these results to rank models overall.
Qwen3.8 Max
Alibaba
Reported by AlibabaMuse Spark Open
Meta
Reported by MetaMuse Spark 1.1
Meta
Reported by Meta11 modelsMultimodal & GroundedCurrentDisplay onlyUpdated November 13, 2026
Benchmark score table (11 models)
All 11 results on this page were reported by the labs that built the models.
| Rank | Model / configuration | Score | Who ran it | Parameters (B) | Open / closed |
|---|---|---|---|---|---|
| 1 | Qwen3.8 MaxAlibabaReported by Alibaba | 82.0% | Reported by Alibaba | Not reported | Open |
| 2 | Muse Spark OpenMetaReported by Meta | 80.2% | Reported by Meta | Not reported | Open |
| 3 | Muse Spark 1.1MetaReported by Meta | 76.3% | Reported by Meta | Not reported | Closed |
| 4 | Seed 2.1 ProByteDanceReported by ByteDance | 73.7% | Reported by ByteDance | Not reported | Closed |
| 5 | Inkling 2 SmallThinking Machines LabReported by Thinking Machines Lab | 68.6% | Reported by Thinking Machines Lab | Not reported | Closed |
| 6 | Qwen3.8-27BAlibabaReported by Alibaba | 65.7% | Reported by Alibaba | Not reported | Open |
| 7 | Seed 2.1 TurboByteDanceReported by ByteDance | 62.9% | Reported by ByteDance | Not reported | Closed |
| 8 | GLM-5.3-FlashZ.AIReported by Z.AI | 53.4% | Reported by Z.AI | Not reported | Open |
| 9 | dots3-note PreviewDots StudioReported by Dots Studio | 50.0% | Reported by Dots Studio | Not reported | Open |
| 10 | Nemotron 3.5 Super VL 120B A12BNVIDIAReported by NVIDIA | 29.6% | Reported by NVIDIA | Not reported | Open |
| 11 | Nemotron 3.5 Super VL 120B A12B NVFP4NVIDIAReported by NVIDIA | 29.5% | Reported by NVIDIA | Not reported | Open |
- Lab-reported
- Benchmark-run
- Independent
Among the reported BabyVision rows, Qwen3.8 Max is first at 82.0%. The third row is 5.7 points behind. The broader top-10 range is 52.4 points, so the table still separates the published systems.
11 models have been evaluated on BabyVision. The benchmark falls in the Multimodal & Grounded category. BabyVision is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.
About BabyVision
Year
2026
Tasks
Visual perception tasks
Format
Multimodal visual reasoning
Difficulty
Fine-grained visual perception
BenchLM tracks BabyVision as a display-only multimodal perception benchmark when providers publish exact comparison values. It complements chart-heavy benchmarks such as CharXiv by emphasizing grounded visual understanding.
Freshness and provenance
Version
BabyVision 2026
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
Questions
What does BabyVision measure?
A multimodal benchmark for fine-grained visual perception and grounded reasoning tasks.
Which model scores highest on BabyVision?
Qwen3.8 Max by Alibaba currently leads with a score of 82.0% on BabyVision.
How many models are evaluated on BabyVision?
11 AI models have published results on BabyVision in the BenchLM catalog.
Compare top models on BabyVision
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 7,000+ subscribers.
One email each week. Unsubscribe anytime.