v0.1.18

What's Changed

• Add option (autoselect_llm) to automatically select the LLM by @dilithjay in #118
• Add support for reference highlighting by @dilithjay in #127
• Parse to latex by @Vaishnav2804 in #120
• Add support for bounding box extraction in images and LLM_PARSE via PaddleOCR by @dilithjay in #128
• Fix injection through doc path by @dilithjay in #130

Full Changelog: v0.1.17...v0.1.18

v0.1.17

What's Changed

• Add support for dataclass in parse_with_schema function
• Upgrade Anthropic version
• Check for title attribute in web pages
• Fix arxiv URL parsing
• Handle invalid bytes when text parsing

Full Changelog: v0.1.16...v0.1.17

v0.1.16

What's Changed

• Set thinking budget to fix gemini-2.5-pro thinking for too long
• Update benchmark code and add more metrics
• Add support for SmolDocling
• Add support for Mistral OCR

Full Changelog: v0.1.15...v0.1.16

v0.1.15

What's Changed

• Fallback to different parser in AUTO mode or STATIC mode
• Add Gemini support to parse pdf with schema
• Add support for Anthropic models
• Update benchmark

Full Changelog: v0.1.14...v0.1.15

v0.1.14

What's Changed

• Add support for Fireworks API
• Add support for matching data in a document to a pre-defined schema or template

Full Changelog: v0.1.13...v0.1.14

v0.1.13

What's Changed

• STATIC_PARSE improvements
  • Horizontal line detection
  • Strikethrough text detection
  • Email address formatting
  • Improved heading level detection
  • Monospace font detection
  • Indentation detection

Full Changelog: v0.1.12...v0.1.13

v0.1.12

What's Changed

• Add support for OpenRouter models
• Return token cost when cost mapping is provided
• Set default router_priority to speed
• Add support for custom prompts
• Add support for parsing Excel and PowerPoint files

Full Changelog: v0.1.11...v0.1.12

v0.1.11.post1

What's Changed

Patch release:

• Fix seg fault when PyQT app reinitialization
• Update Google Colab link in usage to match updated link in badge
• Add Code of conduct

Full Changelog: v0.1.11...v0.1.11.post1

v0.1.11

What's Changed

• Added priority setting to AUTO routing
• Added more models to benchmark
• Set default parse_type to AUTO
• Set default LLM to Gemini 2.0 Flash
• Updated benchmark script to aggregate over multiple runs
• Fixed incorrect title for webpage when as_pdf=True

Full Changelog: v0.1.10...v0.1.11

v0.1.10

What's Changed

• Add param to specify page numbers to parse
• Fix errors caused by empty token_usage

Full Changelog: v0.1.9...v0.1.10