SpeechGen

SpeechGen is an online text to speech service that converts typed or pasted text, and uploaded PDF and DOCX files, into spoken audio. The site lists more than 5,000 voices across about 150 languages, with regional accents and speaking styles such as cheerful, sad and whisper. An SSML editor controls pauses, emphasis, pitch, rate and volume, and multi speaker dialogues can be built in a single file. Output downloads as MP3, WAV, OGG, FLAC and several other formats, and background music can be mixed in. The service also transcribes audio to text. Transcription accepts files up to 1 GB or three hours, supports speaker labeling, word level timestamps and SRT and VTT subtitle export, and takes MP3, WAV, YouTube links and video. An API is included for automating work at scale. It suits video creators, podcasters, e learning producers and anyone who needs narration without recording it. Pricing is pay as you go with no subscription. Credits are bought in packs: 25,000 credits for $4.99, 65,000 for $9.99, 200,000 for $24.99, 500,000 for $49.99, 2,000,000 for $149.99 and 10,000,000 for $599.99. Credits last one year and are spent at different rates by voice tier, with Standard voices covering two characters per credit, Pro one and HD half a character. Every pack includes all features and a commercial license. New visitors can try 1,000 characters free without an account.