SpeechSuper provides pronunciation and speaking-assessment APIs and SDKs for language-learning and oral-practice products. It uses deep learning models to analyze speech data across dimensions such as pronunciation, fluency, grammar, and vocabulary. Its primary audience is developers and edtech companies, rather than individual learners using a standalone app.
Its API supports scripted pronunciation assessment in 8 languages: English, Mandarin Chinese, German, French, Spanish, Korean, Japanese, and Russian. It can handle words/single characters, short texts of 2-200 words or characters, and long texts of more than 200 words or characters. For English, it also supports phoneme-sequence assessment, semi-scripted speaking assessment for PTE, and open speaking assessment. The open assessment can output transcription, overall score, fluency, grammar, vocabulary, and pronunciation. Mandarin capabilities are more granular, covering initial/final sounds, tone accuracy, sentence-final intonation, and more. The mobile SDK supports iOS and Android, but only for English and Mandarin, with an emphasis on offline use, local data, and unlimited requests.
The API is billed by request volume rather than audio duration. Pay-as-you-go has a monthly minimum of $20 and is billed quarterly. Scripted assessment ranges from $0.004 to $0.008/request, semi-scripted English costs $0.025/request, and open English costs $0.030-$0.035/request. Prepaid plans start at $500 and include tiered discounts, with custom pricing above $5000. The SDK is priced as an annual per-device fee, starting from 2,000 devices, at roughly $1 to $0.90/SDK/year. The free trial requires a business email; according to the official site, an expert will contact you within 1 business day, so it is not a fully self-service signup.
Its strengths are the fine-grained assessment levels, down to phoneme, character, word, sentence, and paragraph level, making it suitable for pronunciation correction, read-aloud scoring, and spoken test practice. The API and SDK offerings are clearly defined, and pricing is relatively transparent. Limitations include the lack of disclosed accuracy benchmarks, SLA, compliance certifications, data retention periods, and other information commonly needed for enterprise procurement. Open speaking assessment is only explicitly available for English, and the SDK supports fewer languages than the API.
SpeechSuper is suitable for language-learning apps, online education platforms, PTE/IELTS preparation products, and enterprise development teams that need to embed speaking-scoring capabilities. Access from mainland China and supported payment methods are not specified in the text, so china_access can only be classified as unknown. If network access or procurement is restricted, alternatives to compare include Azure AI Speech, Google/AWS speech services, and domestic options such as iFLYTEK speech assessment.
β 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 speechsuper.com official site.
speechsuper.com is an Unknown AI Apps provider. TG4G tracks its product information, an overall rating of 8.0/10, and a China-accessibility score of China direct-connect friendly. Click "Visit Official Site" to reach speechsuper.com directly.