Rime Speech AI Software
About Rime
Rime was founded in San Francisco in 2022 by Lily Clifford, Brooke Larson, and Ares Geovanos. Clifford left a Stanford NLP PhD to start it, Larson is a PhD linguist who worked on Amazon Alexa, and Geovanos is a Stanford engineer. The company describes its mission as making voice AI sound and feel human on the phone.
The approach that sets it apart is data. Rime built a recording studio in San Francisco and captured its own set of spontaneous, full duplex conversations, including interruptions and disfluencies, rather than training on audio scraped from the web. It pairs that with a phoneme based design and pronunciation controls so customers can fix how a brand name or medication sounds without retraining a model.
The product line now covers the whole call. Mist v3 is a low latency text-to-speech model and Coda v2 is a more natural one, with more than 600 voices across many languages between them. Rime STT handles transcription of 8 kHz phone audio. Prism, launched in 2026, is a speech-to-speech model that reasons over the caller's audio, produces text before it speaks so compliance checks can run, and supports tool calls mid-call.
Rime works with voice agent platforms such as LiveKit and Pipecat and with enterprises directly. Dialpad was a design partner on Prism. On July 15, 2026 Rime announced a $24 million Series A led by M13, with Twilio Ventures, Corazon Capital, and Unusual Ventures, and hired Rafael Valle from Meta as chief scientist. TechCrunch reported a team of 35 at the time.
Rime Features
1. Prism speech-to-speech model with a text checkpoint before audio for guardrails
2. Mist v3 text-to-speech with sub-100 millisecond time to first byte
3. Coda v2 text-to-speech for the most natural sounding voices
4. Rime STT speech-to-text built for live phone audio
5. More than 600 voices with control over accent, pace, and tone
6. Deterministic pronunciation and custom lexicons for names, brands, and alphanumerics
7. Tool calls and interruption handling during live calls
8. Integrations with LiveKit, Pipecat, and other voice agent frameworks
9. Cloud, VPC, or on-premises deployment, with a HIPAA BAA and SOC 2 Type II report
Rime Pricing
Rime publishes its pricing. The Free plan includes $50 of usage each month with no card required. Flex is pay as you go with no monthly fee: Prism speech-to-speech costs $0.080 per minute, Mist v3 text-to-speech $0.030 per minute, Coda v2 $0.050 per minute, and Rime STT $0.002 per minute. Scale costs $100 a month, includes $100 of credit that rolls over up to $200, and takes 20 percent off every rate, so Prism drops to $0.064 per minute. Enterprise is custom with annual billing, volume rates, cloud, VPC, or on-premises deployment, custom voices, fine-tuning, an SLA, a HIPAA BAA, and a SOC 2 Type II report.
Rime Customers
Dialpad; Mayo Clinic; Upstart; Asurion; ConverseNow; Goodcall; Twilio; Trillet AI; SigmaMind AI; Attune
Other Speech AI Providers


