Fixed - Segmentation fault when PyQT app is reinitialized
0.1.11
Added - Priority setting to AUTO routing - More models to benchmark
Changed - 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 when `as_pdf=True`
0.1.10
Added - Parameter to specify page numbers for parsing
Fixed - Errors caused by empty token_usage
0.1.9
Added - Parameters to specify intermediate PDF save path when `as_pdf=True`. - Return `token_uage` and `pdf_path` with `parse()` output where applicable
Changed - Switched back to together Python client - Improved `parse()` function return format to be a dictionary.
0.1.8.post1
Added * Documentation
Changed * Specify headers for Playwright web page retrieval
0.1.8
Added * Retry and error handling for LLM_PARSE
Changed * Remove together Python client dependency and use REST API calls instead