Docbench - Slides: leaderboard

Docbench in Gezel: fixed document-task adapters, parser/repair/retry behavior, native checks plus Claude Opus 5.5 rubric judge. Unusable outputs score zero. 32, 768 context except Apple 8, 192. Mostly Q4 is not affirmative precision evidence for each row; preserve explicit Q2/Q4 only. CLI/API labels are run configuration, not distinct public models. Native Apple Foundation Models row retained in pinned capture but not emitted until exact checkpoint identity is established. 102 public cases, 142 outputs; some models have three repetitions. Overall family weights25/25/20/10/15/5; 95% range is uncertainty, not another observation. Retain each native family grader mixture. Metric: Native task quality score (%). Source: https://bendyline.com/posts/docbench.html

Metric: Native task quality score (%). Source: bendyline.com. Saturation forecast: Around February 2027. 29 models tracked. Updated 17 Nov 2026.

Top models

#ModelScoreOverall rank
1GPT-6.1 Sol94#12
2Claude Opus 5.592#7
3Claude Sonnet 4.691#173
4Claude Sonnet 5.591#21
5Qwen 3.6 35B A3B89#484
6Gemma 4 26B A4B89#581
7Qwen 3.8 27B (Q4)89
8Qwen 3.8 27B (Q2)89
9GPT-5.6 Sol88#33
10Claude Haiku 4.587#580
11Gemma 4 E4B86#2142
12Gemma 4 12B86#931
13Ornith-1.5-9B86#843
14Qwen 3.5 4B85#1227
15Qwen 3.5 9B84#822

No result here: #1 Claude Fable 5.2, #2 GPT-6.2 Astra, #3 Gemini 4 Carbon.

Interactive version: theaggregate.ai/benchmark?slug=docbench-slides · How It Works · Data refreshed daily, snapshot 2026-11-17.