Benchmark Almanac

NEST

Qwen3.5-27B 76% leads since November 2026
Not in model cards

Models interpret children's gaze, gestures, and physical contact in social videos. The benchmark checks classification and choice questions and separately grades written descriptions.

Released
November 2026
Kind
Video understanding
Level
Everyday
Scored
Checked automatically
Answers
Multiple choice
Tasks
All public
Measures
Capability
Language
English
In and out
Video, Text → Text
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
In views
Vision
Strongest model tested
GPT-5 mini #92 on the capability index
Last model added
March 2026
Scores refreshed
November 20th, 2026
17
Entries on its own leaderboard, 12 models we track

Qwen3.5-27B leads with 76% among models released by March 2026 Qwen3.5-27B leads with 76% among models released by March 2026

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 March 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 Qwen3.5-27B 76%
  2. 2 Qwen3.5-9B 73.9%
  3. 3 GPT-5 mini 73.4%
  4. 4 Qwen3.5-4B 69.7%
  5. 5 Gemini 3.1 Flash-Lite 69.7%
  6. 6 Molmo2-8B 69.6%
  7. 7 MiniCPM-V 4.5 67.6%
  8. 8 Qwen3-VL 32B 66.4%
  9. 9 openPangu-VL 7B 63.4%
  10. 10 LLaVA-OneVision 7B 59.6%
  11. 11 Qwen3-VL 8B 59.1%
  12. 12 InternVL3.5-38B 57.4%
  13. 13 Qwen3-VL 4B 55.5%
  14. 14 Keye-VL-1.5-8B 54.8%
  15. 15 InternVL3.5-8B 46.1%
  16. 16 LLaVA-NeXT-Video 7B 44%
  17. 17 InternVL3.5-4B 41.9%

Source: Official NEST leaderboard

#EntryScoreSettingDateSource
1 Qwen3.5-27B 76% Feb 2026 Official leaderboard
2 Qwen3.5-9B 73.9% Feb 2026 Official leaderboard
3 GPT-5 mini 73.4% Aug 2025 Official leaderboard
4 Qwen3.5-4B 69.7% Feb 2026 Official leaderboard
5 Gemini 3.1 Flash-Lite 69.7% Mar 2026 Official leaderboard
6 Molmo2-8B 69.6% Dec 2025 Official leaderboard
7 MiniCPM-V 4.5 67.6% Aug 2025 Official leaderboard
8 Qwen3-VL 32B 32B 66.4% Official leaderboard
9 openPangu-VL 7B 7B 63.4% Official leaderboard
10 LLaVA-OneVision 7B 59.6% Aug 2024 Official leaderboard
11 Qwen3-VL 8B 8B 59.1% Official leaderboard
12 InternVL3.5-38B 57.4% Aug 2025 Official leaderboard
13 Qwen3-VL 4B 4B 55.5% Official leaderboard
14 Keye-VL-1.5-8B 54.8% Aug 2025 Official leaderboard
15 InternVL3.5-8B 46.1% Aug 2025 Official leaderboard
16 LLaVA-NeXT-Video 7B 7B 44% Official leaderboard
17 InternVL3.5-4B 41.9% Aug 2025 Official leaderboard
#ModelScoreSettingDateSource
1 Qwen3.5-27B 76% Feb 2026 Official leaderboard
2 Qwen3.5-9B 73.9% Feb 2026 Official leaderboard
3 GPT-5 mini 73.4% Aug 2025 Official leaderboard
4 Qwen3.5-4B 69.7% Feb 2026 Official leaderboard
5 Gemini 3.1 Flash-Lite 69.7% Mar 2026 Official leaderboard
6 Molmo2-8B 69.6% Dec 2025 Official leaderboard
7 MiniCPM-V 4.5 67.6% Aug 2025 Official leaderboard
8 LLaVA-OneVision 7B 59.6% Aug 2024 Official leaderboard
9 InternVL3.5-38B 57.4% Aug 2025 Official leaderboard
10 Keye-VL-1.5-8B 54.8% Aug 2025 Official leaderboard
11 InternVL3.5-8B 46.1% Aug 2025 Official leaderboard
12 InternVL3.5-4B 41.9% Aug 2025 Official leaderboard

Rankings by part

We read 3 parts from the benchmark’s own leaderboard: Contextual, Behavioral and Interpersonal. Each ranks models on one part of the test. Qwen3.5-27B leads 2 of them, and GPT-5 mini leads the other one. Arrows show how far a model moved from its overall rank among the models we track.

Contextual

12 models we track have a score here.

  1. 1 Qwen3.5-27B 88.5%
  2. 2 Qwen3.5-4B ↑2 85.5%
  3. 3 Qwen3.5-9B ↓1 85.2%
  4. 4 Gemini 3.1 Flash-Lite ↑1 83.2%
  5. 5 GPT-5 mini ↓2 82.6%
  6. 6 MiniCPM-V 4.5 ↑1 82.6%
  7. 7 Molmo2-8B ↓1 81.2%
  8. 8 LLaVA-OneVision 7B 72.7%
  9. 9 InternVL3.5-38B 65.4%
  10. 10 Keye-VL-1.5-8B 60.8%

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

Qwen
Own leaderboard 1 leader

Source: Official NEST leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On the benchmark’s own leaderboard Best so far
0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) Qwen3.5-27B GPT-5 mini LLaVA-OneVision 7B 0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) Qwen3.5-27B LLaVA-OneVision 7B

Source: Official NEST 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

All 695 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 8 months ago, in March 2026, for Gemini 3.1 Flash-Lite. 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 (17 entries).
  • Pass Models scored 12 models we track, plus 5 other entries on its leaderboard.
  • Warning Newest score The newest came 8 months ago, in March 2026, for Gemini 3.1 Flash-Lite.
  • Note Frontier models None of the top 10 has a score. The strongest tested is GPT-5 mini (#92).
  • Pass Headroom The best is 76%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on laubravo.github.io: 12 rows for models we track and 5 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.

How we check