Deepgram Expands Nova-3 with Spanish, French, and Portuguese Support
Blog post from Deepgram
Deepgram has expanded its Nova-3 speech-to-text model to support Spanish, French, and Portuguese variants, including Latin American Spanish, Canadian French, and Brazilian and European Portuguese, following earlier additions of several northern European languages. The company says the expansion addresses linguistic challenges such as dialect variation, connected speech, nasalization, and differing phonetics, enabling uses in contact centers, healthcare, banking, legal transcription, and customer support across global markets. Nova-3 includes features such as keyterm prompting, real-time scalability, and transcription performance in noisy environments, with reported relative streaming word error rate reductions versus Nova-2 of 24.35% for Portuguese, 21.40% for Spanish, and 14.26% for French. Users can access the languages by specifying the appropriate language parameter in Nova-3 API requests.
Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.