Dimension scores are derived from public data and fields; weighted into the composite. Reference only.
SpeechText.AI is an AI-powered audio and video transcription service that offers both a web-based transcription tool and a Speech-to-Text API. It supports uploads in common audio and video formats, automatically generates punctuated transcripts, and can export to TXT, PDF, DOCX, SRT/VTT, and other formats. It is positioned for use cases such as interviews, meetings, podcasts, legal, medical, customer support, and developer integrations.
The product is built around deep neural network speech recognition and highlights its “domain models”: users can choose industry domains such as legal, medical, finance, HR, and customer support to improve recognition of specialized terminology. Features include support for 50+ languages, non-native accent support, speaker identification, automatic punctuation, online editing, audio search, summaries, and keyword highlighting. The main content explicitly supports Mandarin Chinese, but does not clarify Chinese dialect support, Simplified/Traditional Chinese handling, or details of a Chinese-language interface. Its accuracy claims include a 3.8% word error rate on the LibriSpeech English dataset, while the German page also cites ranges such as 93.8%–96.2%; however, the site also acknowledges that noise, overlapping speech, and recording quality can affect results.
The web version uses pay-as-you-go pricing with no monthly fee: from $10/180 minutes to $99/2000 minutes, with file size limits ranging from 30MB to 1GB. The API is sold as a monthly subscription: from $49/2700 minutes to $399/33250 minutes, or roughly $0.018–$0.012 per minute, and it offers a free API Key. The exact number of free trial minutes and its limitations are not clearly stated in the main content.
Its strengths are broad language and format coverage, and its specialized domain models are useful for terminology-heavy content. API examples cover Python, cURL, PHP, and Java, and support binary uploads, public URLs, SRT output, summaries, and more, making integration relatively straightforward. On privacy, the site states that it uses European/French servers, is GDPR-compliant, uses encrypted transmission, and allows files to be deleted; the API FAQ also says files are deleted immediately after transcription is completed. The downsides are that its accuracy figures mainly come from website claims or specific dataset results, so performance on complex real-world recordings still needs testing; information on free usage allowance, SLA, payment methods, and enterprise support tiers is also limited.
SpeechText.AI is suitable for podcast creators, journalists, research interviewers, meeting-minutes teams, professional users in legal/medical fields, and developers who need a low-cost speech transcription API. The main content does not provide information on access status from mainland China, nor does it specify payment methods. If network access or payment is restricted, users may want to compare it with Google Speech-to-Text, Amazon Transcribe, Azure Speech, Whisper, or local speech recognition services.
⚠ This review is compiled from public sources and does not constitute a purchase recommendation. Verify all facts on the vendor's official site. Verify on speechtext.ai official site.
speechtext.ai is an Unknown AI Apps provider. TG4G tracks its product information, with monthly pricing from $10.00, an overall rating of 8.0/10, and a China-accessibility score of China direct-connect friendly. Click "Visit Official Site" to reach speechtext.ai directly.