Voice, Speech & Realtime AI

Rime raises $24M Series A to advance phoneme-based speech-to-speech AI

Source: TheAIInsider · Jul 27, 2026

Rime, a San Francisco voice AI startup founded in 2022, has closed a $24 million Series A led by M13 Ventures to advance its move toward integrated speech-to-speech modeling. The round is modest by 2026 voice-AI standards — peers in the category have raised well over $100 million in comparable stages — but Rime's pitch is deliberately narrower: proprietary conversational data collection paired with a phoneme-based architecture, rather than competing head-on for general-purpose voice-platform share.

The phoneme-based approach is the technical differentiator worth noting. Where many voice AI systems still rely on cascaded pipelines — speech-to-text, a language model, then text-to-speech — Rime is building toward a more integrated speech-to-speech architecture that reasons and responds closer to the audio signal itself. That approach typically trades some general-purpose flexibility for lower latency and more natural conversational timing, which matters most in the kind of live, high-stakes conversational settings Rime's client list points toward.

Mayo Clinic and Dialpad as named clients signal Rime is already operating in demanding, real-time conversational environments — healthcare and business communications — rather than staying in a pure-research or demo phase. Both use cases put a premium on accurate, low-latency conversational handling, which lines up with the phoneme-based bet.

Key Points

• $24M Series A led by M13 Ventures for the SF-based voice AI startup, founded 2022
• Differentiates via proprietary conversational data collection and a phoneme-based architecture
• Moving toward an integrated speech-to-speech model rather than a cascaded pipeline
• Named enterprise clients include Mayo Clinic and Dialpad