Releases: letssummarize/api
Releases · letssummarize/api
api: v0.3.0
0.3.0 (2025-03-28)
Features
- add abillity to add multible allowed origins (172a497)
- add ApiKeyGuard for external integrations, improve options, and clean up code (1fc48bb)
- add audio option for summary, implement automatic cleanup for audio files (45dcefb)
- add audioFilePath to response (f64319c)
- add customization text support to summarization prompt (5817228)
- add language selection for summarization (Arabic/English) (f6342cb)
- add speed option (fast/slow) for YouTube summarization (0178c5e)
- add support for Fast-Whisper as an option and integrate with the Fast-Whisper Python API (70e2730)
- add support for Gemini model for summarization (5aae38e)
- added file summarization (e9d3108)
- added video summarazation (d64dd34)
- api: increase max tokens for models (3657265)
- enable CORS (1e46008)
- enable S3 storage for all files not only the TTS generated files (6fe7907)
- enforce max length for custom instructions and prioritize them in summarization (4cd8452)
- enhance API key guard to allow API users of Gemini and Fast-Whisper without an API key (c14da72)
- improve prompts content (51df964)
- improve response formatting and add video metadata (045fb8c)
- init trascribe api using faster-whisper (WIP) (869548b)
- storage: enable optional S3 upload for TTS-1 generated audio files (87bb148)
- support deepseek model in request options (ed66c2d)
- use text language for summary if 'lang' is default or missing (7f4442f)
- validate options by setting defaults values and improve prompts (b0a2c63)
- video summarization: prioritize youtube-transcript api before downloading audio (dd4ee88)
Bug Fixes
- add YouTube session cookies for authentication (89f8d39)
- added referer and user-agent headers to ytDlpExec to avoid request blocking (0f9291c)
- audio should be always available when openai default key exists (3bf288d)
- authenticate yt-dlp with GVS using PO token (f95f77e)
- dockerfile: update CMD path for transcribe_api (8371112)
- download: force generic extractor to bypass YouTube restrictions (09c2e53)
- enable global validation and insure SummarizationOptionsDto is validated in controller (96f2f78)
- ensure yt-dlp-exec can download yt-dlp binary (fcda3e2)
- error message in summarizeYouTubeVideo (ab14827)
- exclude language from prompt when format is custom and language is default (263545b)
- file: file summarization was broken down while refactoring (2631708)
- fix file type validation (18d9949)
- improve error handling by differentiating between missing transcripts and invalid API keys (1c36ba3)
- improve error handling for summarization APIs (a3aacba)
- return response if no YouTube transcript, restrict DeepSeek users from audio & slow mode (99f818f)
- send correct tts file path when slow speed is provided (aa0ab6f)
- slow speed was not working with s3 storage (a87580e)
- transcribe api: FastAPI server was not running (WIP) (98f0b20)
- transcribe api: resolve loading model and transcribe audio issues (WIP) (fa31c3b)
- transcribe-api: optimize model loading and caching for better performance (WIP) (1062be9)
api: v0.2.0
0.2.0 (2025-03-18)
Features
- add ApiKeyGuard for external integrations, improve options, and clean up code (1fc48bb)
- add audio option for summary, implement automatic cleanup for audio files (45dcefb)
- add audioFilePath to response (f64319c)
- add speed option (fast/slow) for YouTube summarization (0178c5e)
- added file summarization (e9d3108)
- added video summarazation (d64dd34)
- enable CORS (1e46008)
- improve response formatting and add video metadata (045fb8c)
- init trascribe api using faster-whisper (WIP) (869548b)
- storage: enable optional S3 upload for TTS-1 generated audio files (87bb148)
- support deepseek model in request options (ed66c2d)
- validate options by setting defaults values and improve prompts (b0a2c63)
- video summarization: prioritize youtube-transcript api before downloading audio (dd4ee88)
Bug Fixes
- dockerfile: update CMD path for transcribe_api (8371112)
- fix file type validation (18d9949)
- transcribe api: FastAPI server was not running (WIP) (98f0b20)
- transcribe api: resolve loading model and transcribe audio issues (WIP) (fa31c3b)
- transcribe-api: optimize model loading and caching for better performance (WIP) (1062be9)