Changelog - API & SDK Version History

• EPUB e-book, SVG vector graphics and EML email compression
• Text family expansion: 19 new extensions including Markdown / LaTeX / YAML / TOML / GeoJSON / KML / XBRL, with tiered zstd/brotli lossless encoding
• Industry & scientific formats: DICOM medical imaging (lossless re-sealing), NIfTI neuroimaging, DXF, LAS/LAZ point clouds, HDF5/NetCDF, KMZ, Shapefile bundle packaging
• TIFF/GeoTIFF native lossless channel: byte-identical pixels, georeferencing and other business tags fully preserved

• Deep Office compression: recursive embedded-object compression + embedded font subsetting + OLE bitmap compression — a real-world ERP spreadsheet sample reaches 11.64x at the ultra level
• OFD engine rework: 2.92x ratio at the ultra level, 4~11x faster than the legacy engine, with signature/attachment whitelist protection
• TIFF large-file processing 2.5x faster (strip-parallel compression + upfront gain estimation)
• Quality floors across the board: SSIM≥0.90 / PSNR≥30dB; no-gain files are automatically kept byte-identical, never larger

• Batch compression endpoint POST /v1/batch/compress
• Usage statistics endpoint GET /v1/usage

• PDF compression speed improved by 30%
• Error responses now include error_detail field for easier troubleshooting

• Added OFD format compression support
• Encrypted compression password parameter

• Python SDK supports async operations

• Large file upload timeout issue

• API officially released
• Supports PDF / image / Office document compression
• Supports curl / Python / Java / Node.js / Go
• SDK v1.0.0 released simultaneously

All currently released API and SDK endpoints are in stable and available status. If any endpoints are planned for deprecation, we will notify at least 90 days in advance and provide migration guides.