VOLTBench - Mean Length Accuracy (200 Sections): leaderboard

Length adherence on VOLTBench (2026), a long-form generation benchmark of chapter-based story and code instructions, at the 200-section target: five generations are sampled per instruction and their mean length is scored as 100 x max(0, 1 - |mean - target| / target). Percent; higher is better.

Metric: Mean length accuracy (%): 100 x max(0, 1 - |mean output length - target length| / target length) over the five sampled generations per instruction, English simple-difficulty instructions, Story (unstructured) and Code Function (structured) chapter-based tasks, 200-section target; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 9 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1LongWriter-8B15.9
2GPT-4o Mini14#1667
3DeepSeek V33.8#729
4DeepSeek R13.4#482
5Qwen 2.5 7B Instruct1.4#2723
6Qwen 2.5 1.5B Instruct0.9#7470
7Llama 3.1 8B Instruct0.8#4600
8Claude 3.5 Sonnet0.5#752
9Falcon3-Mamba-7B-Instruct0.4

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

Interactive version: theaggregate.ai/benchmark?slug=voltbench-mean-length-accuracy-200-sections · How It Works · Data refreshed daily, snapshot 2026-11-21.