As of: 2026-11-22 13:32 CET. Researched and written by AI agents (Claude Opus 5.5 in Claude Code), with a human editor. How we work: https://postcutoff.com/about/ Canonical page: https://postcutoff.com/m/teleocr/ # TeleOCR (ex NaviDC-OCR) TeleAI, TeleOCR family. Status: current. Type: multimodal model. Released: 2026-08-17. Training cutoff: not published. ## How to call it | Provider | Model id | Endpoint | Docs | |---|---|---|---| | Hugging Face | | | https://huggingface.co/XingChen-AGI/TeleOCR | | GitHub | | | https://github.com/caipeng328/TeleOCR | | Web app | | | https://www.teleai.com.cn/docparse/DocumentParsing | ## Pricing - Price not published. - Last checked: 2026-10-06 ## Specs - Input: image, pdf - Output: text - Context window: not published - Max output: not published - Open weights: yes - Licence: apache-2.0 ## What it has not seen - No published training cutoff. Counted from its release (2026-08-17): at least 1,834 AI events since (424 major, 78 historic), as of 2026-11-22. - Everything since July 2026: https://postcutoff.com/since/2026-07/ ## What stands out - Camera-captured document parsing without dewarping: One ~1.4B-parameter Qwen2.5-VL-based model parses both digital and photographed or curved documents directly, with no separate rectification step (DocUNet/DIR300 demos). Source: https://huggingface.co/XingChen-AGI/TeleOCR - Dr.DocBench Challenge lead (found after launch): Self-reported 67.96 overall on the EMNLP 2026 Dr.DocBench Challenge vs MinerU 2.5 Pro 62.26 and PaddleOCR-VL 1.6 55.11. Source: https://huggingface.co/XingChen-AGI/TeleOCR ## Notes Released Aug 17, 2026 as NaviDC-OCR (StarDoc-AI), renamed TeleOCR on Sept 10. ~1,280 HF likes and ~41k downloads by Oct 6, 2026; GGUF/llama.cpp ports exist. Benchmarks are self-reported. The README links TeleAI's (China Telecom) document-parsing web app; the HF org is XingChen-AGI. Lightweight open document-parsing VLM. Technical report: arXiv 2608.12898. Model page: https://postcutoff.com/m/teleocr/