Gradium is a TTS / Voice Synthesis company founded in 2025 and based in Paris, France. It has raised $100M in total funding, most recently a Seed in 2026.
| Date | Stage | Amount | Valuation | Lead investors |
|---|---|---|---|---|
| Jul 9, 2026 | Seed | $100M | — | FirstMark Capital, Eurazeo |
No articles ingested yet for Gradium. Once the hourly news pipeline is live, every article the classifier tags as mentioning this company appears here with its one-line AI summary and sentiment.
Gradium's real-time text-to-speech API turns text into speech in English, French, German, Spanish and Portuguese, with instant and professional voice cloning and built-in normalization for phone numbers, dates, codes and URLs. Voice Design, launched Sept. 7, 2026, creates a new voice from a text prompt, and Gradium added more than 1,000 voices on Oct. 7, 2026. Through LiveKit Inference it reports time to first audio under 250 milliseconds.
Gradium's real-time transcription model emits predictions every 80 milliseconds and includes semantic turn detection, which judges whether a speaker has finished talking. Voice agents use it to decide when to respond without cutting users off.
Gradium Translate, launched June 24, 2026, performs real-time speech-to-speech and speech-to-text translation across English, French, German, Spanish and Portuguese. It is billed per second of audio through the same credit plans as Gradium's other APIs.
Phonon is an open-weight, on-device text-to-speech model released under the Apache 2.0 license on Oct. 21, 2026, in 31-million and 40-million-parameter versions. It generates speech locally on phones, laptops, embedded hardware, browsers and NVIDIA Jetson platforms, in five languages. Hugging Face, Arm and Qualcomm were among its launch partners.
We don't have a live feed for this company's ATS. Their careers page has every open role.
View all careers ↗