MMLongBench-Doc
We show this table for reference; we do not rank on it.
A long-document multimodal benchmark for grounded reasoning over extended document contexts.
Benchmark score on MMLongBench-Doc — November 13, 2026
We compile the MMLongBench-Doc rows from provider self-reports and secondary reports. Nemotron 3.5 Super VL 120B A12B leads the table at 62.8%, followed by Nemotron 3.5 Super VL 120B A12B NVFP4 (62.8%) and Qwen3.5 397B (61.5%). We do not use these results to rank models overall.
Nemotron 3.5 Super VL 120B A12B
NVIDIA
Reported by NVIDIANemotron 3.5 Super VL 120B A12B NVFP4
NVIDIA
Reported by NVIDIAQwen3.5 397B
Alibaba
Reported by Alibaba6 modelsMultimodal & GroundedCurrentDisplay onlyUpdated November 13, 2026
Benchmark score table (6 models)
None of the 6 results on this page has a documented run by the benchmark or an independent evaluator: 4 reported by the labs that built the models and 2 reported in another publisher’s document.
| Rank | Model / configuration | Score | Who ran it | Parameters (B) | Open / closed |
|---|---|---|---|---|---|
| 1 | Nemotron 3.5 Super VL 120B A12BNVIDIAReported by NVIDIA | 62.8% | Reported by NVIDIA | Not reported | Open |
| 2 | Nemotron 3.5 Super VL 120B A12B NVFP4NVIDIAReported by NVIDIA | 62.8% | Reported by NVIDIA | Not reported | Open |
| 3 | Qwen3.5 397BAlibabaReported by Alibaba | 61.5% | Reported by Alibaba | Not reported | Open |
| 4 | Gemini 3 ProGoogleAnother lab’s comparison table | 60.5% | Another lab’s comparison table | Not reported | Closed |
| 5 | Kimi K2.5Moonshot AIAnother lab’s comparison table | 58.5% | Another lab’s comparison table | Not reported | Open |
| 6 | Nemotron 3 Nano Omni 30B A3BNVIDIAReported by NVIDIA | 57.5% | Reported by NVIDIA | Not reported | Open |
- Lab-reported
- Benchmark-run
- Independent
Among the reported MMLongBench-Doc rows, Nemotron 3.5 Super VL 120B A12B is first at 62.8%. The third row is 1.3 points behind. The broader top-10 range is 5.3 points, so many of the published results sit in a relatively narrow band.
6 models have been evaluated on MMLongBench-Doc. The benchmark falls in the Multimodal & Grounded category. MMLongBench-Doc is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.
About MMLongBench-Doc
Year
2026
Tasks
Long document understanding
Format
Document-grounded reasoning
Difficulty
Long-context document reasoning
MMLongBench-Doc is designed to test whether a model can maintain grounded understanding across large document contexts rather than only short OCR-style snippets.
Freshness and provenance
Version
MMLongBench-Doc 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 MMLongBench-Doc measure?
A long-document multimodal benchmark for grounded reasoning over extended document contexts.
Which model scores highest on MMLongBench-Doc?
Nemotron 3.5 Super VL 120B A12B by NVIDIA currently leads with a score of 62.8% on MMLongBench-Doc.
How many models are evaluated on MMLongBench-Doc?
6 AI models have published results on MMLongBench-Doc in the BenchLM catalog.
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.