evolcSign in

Deepslate

Speech-to-speech voice AI model for developers and enterprises building phone and browser voice agents

Highlights

  • Hiring5 open roles
  • New on EvolcAdded 2026-10-17
  • Remote-friendly4 of 5 open roles remote or hybrid
  • Salary transparent5 of 5 open roles list a salary

About Deepslate

Deepslate offers a native speech-to-speech voice AI model that takes audio input and produces audio output directly, without separate speech-to-text, LLM and text-to-speech components. It is accessed through WebRTC, SIP, WebSocket, Python and Node SDKs, and plugins for LiveKit and Pipecat.

The website targets developers building voice agents, and enterprises such as contact centers, insurers and platforms. It sells pay-as-you-go access from 2 ct per minute under a limited release offer (regular price 10 ct), with €15 of free starter credits, and an Enterprise tier with SLA, Zero Data Retention, self-hosting and dedicated onboarding.

The company states 440 ms response latency, support for 27 languages, EU hosting and GDPR compliance, and says it is ISO 27001 certified. The website announces a €7.7M funding round and lists 42CAP, Alstin Capital and SIVentures as backers. Five open roles are listed (Rust, backend, frontend, model research and social media), with locations given as Berlin and Germany. The founding year and office locations are not stated on the website.

Product: A native speech-to-speech voice AI model accessed via API, SDKs, WebRTC and SIP for building voice agents.

Open jobs (5)

Tech stack

  • Framer