According to a recent LinkedIn post from Together AI, the company is featuring deployment of MiniMax Speech 2.8 Turbo, an enterprise text-to-speech model, on its dedicated infrastructure. The post highlights capabilities aimed at expressive, real-time voice agents, including sound tags for vocal cues such as laughter and breathing.
Meet Samuel – Your Personal Investing Prophet
- Start a conversation with TipRanks’ trusted, data-backed investment intelligence
- Ask Samuel about stocks, your portfolio, or the market and get instant, personalized insights in seconds
The post also points to technical improvements, citing a 60% prosody gain over a prior Speech 2.6 version, sub-250 millisecond end-to-end latency with streaming, and support for more than 40 languages. For investors, this suggests Together AI is deepening its position in voice AI infrastructure, potentially enhancing its appeal to AI-native enterprises building global, latency-sensitive applications.
If adopted by developers and enterprise customers, the expanded speech offering could support higher platform usage and stickiness for Together AI’s services. The emphasis on real-time performance and multilingual reach may also position the company competitively in the growing market for conversational and voice-based AI agents, where infrastructure differentiation can be a key driver of customer acquisition and retention.