Chronicles-OCR - Archaic Text Parsing: leaderboard

Chronicles-OCR (2026) tests vision-language models on historical Chinese writing across the Seven Chinese Scripts, 2,800 expert-annotated images from Oracle Bone to Cursive. This board scores transcription of Oracle Bone, Bronze and Seal texts in their original reading order, as normalized edit similarity from 0 to 1 averaged over the three scripts; higher is better.

Metric: Paragraph-level normalized edit similarity (0-1: one minus the edit distance over the longer sequence) of the transcription in reading order against the reference, on the archaic scripts (Oracle Bone, Bronze and Seal), the paper Average column over the per-script results; higher is better. Source: arxiv.org. Saturation forecast: Not forecast. 26 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Kimi K2.50.22#266
2Qwen 3.5 397B A17BNon-reasoning0.22#264
3Seed 2.0 Pro0.21#222
4Qwen 3.5 35B A3BNon-reasoning0.2#492
5Qwen 3 VL 235B A22B Instruct0.19#532
6Qwen 3 VL 8B Instruct0.18#1029
7Qwen 3 VL 235B A22B (Thinking)0.17#459
8Seed 1.80.17#252
9Gemini 3.1 Pro (Preview)0.15#99
10InternVL3.5-241B-A28B0.13#733
11Qwen 3 VL 2B Instruct0.12#2367
12Qwen 3 VL 8B (Thinking)0.09#910
13Ovis2.6-30B-A3BThinking0.09
14Claude Opus 4.7Thinking0.08#85
15MiMo-V2-Omni0.08#465

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

Interactive version: theaggregate.ai/benchmark?slug=chronicles-ocr-archaic-text-parsing · How It Works · Data refreshed daily, snapshot 2026-11-21.