jina-ocr-v1: Turning PDFs, Scans and Tables into Markdown

Melvin Vivas · X video post · 2026-09-18 · Open on X

Topics: Retrieval-Augmented Generation (RAG) · Level: intermediate

Summary

Jina AI released jina-ocr-v1, a visual document parser post-trained on DeepSeek. It has 3.4B total and 570M active parameters with built-in speculative decoding. It turns PDFs, scans, tables, charts and invoices into clean markdown, which makes it useful for document-ingestion pipelines.

Key points

• jina-ocr-v1 is post-trained on DeepSeek
• 3.4B total parameters, 570M active, with built-in speculative decoding
• Input: PDFs, scans, tables, charts, invoices; output: clean markdown
• Available through Jina Reader via the x-respond-with header
• Useful for parsing visual documents before RAG ingestion

Resources mentioned

• jina-ocr-v1 · tool · huggingface.co · free
  Jina AI's visual document parser that turns PDFs, scans, tables, charts and invoices into markdown.
• Jina Reader · tool · jina.ai · free
  Jina AI's API that converts URLs and documents to LLM-friendly text; it exposes jina-ocr-v1 via `x-respond-with`.
• DeepSeek · tool · deepseek.com · free
  Open-weight large language models from DeepSeek, known for strong reasoning and coding at low cost.
  Also in: Open models now dominate token volume on Vercel AI Gateway (Melvin Vivas on X · notes), Bolt.new Adds Open Models (GLM, DeepSeek, Kimi) via Bolt Forge (Melvin Vivas on X · notes), OpenCode Go: Open Coding Models (DeepSeek, Qwen, Kimi, GLM) for $10/mo (Melvin Vivas on X · notes), Orchestrator + Subagents in opencode with Open Models (Melvin Vivas on X · notes) and 4 more

Try this

• Use jina-ocr-v1 to convert invoices or scanned PDFs to markdown as the ingestion step of a RAG pipeline