- The AI Voice Newsletter
- Posts
- OpenAI Launches gpt‐Realtime Voice Model
OpenAI Launches gpt‐Realtime Voice Model
🔊 Soundcheck
OpenAI debuts gpt‑realtime, turbocharging voice AI.
Vox AI Secures $8.7M for Drive-Thru Voice AI
Voicee opens bidding on foundational voice AI patents
Generative AI voice arrives in Jeep cabins.
Read time: 4 minutes
🔥 Hot Mic
Big moves, deep dives, and standout stories.
OpenAI’s new gpt‑realtime speech model and Realtime API upgrades enable faster, more expressive, tool‑conversation voice agents.
OpenAI just unveiled gpt‑realtime, a speech‑to‑speech model built to power natural, low‑latency voice agents with expressive tone and context awareness. It processes audio directly rather than chaining through text, producing smoother voices and preserving emotional nuance. This release also brings enhanced support within the Realtime API: two new voices (Cedar and Marin), image input, remote tool integration via MCP servers, SIP phone calling, and reusable prompts. The package includes significant benchmark improvements and a 20% price cut, making it an appealing upgrade for real‑time voice applications.
Key Points:
End‑to‑end speech‑to‑speech with no STT/TTS chaining
Two new expressive voices—Cedar (male) and Marin (female)
Significant benchmark gains: Big Bench Audio, MultiChallenge, ComplexFuncBench
Realtime API now supports MCP tools, image inputs, SIP calling
Takeaway: gpt‑realtime reshapes voice AI by combining speed, expressiveness, tool integration and cost savings into a single, more human‑like system—making real‑time conversational agents more capable and accessible.
Vox AI raised $8.7 million seed to expand its autonomous, multilingual voice assistant in QSR operations.
Vox AI, founded in October 2023, builds voice AI tailored for drive-thrus and quick-service restaurant operations. Its platform autonomously handles drive-thru and mobile orders in over 90 languages, fits into existing systems, and assists staff with real‑time alerts and guidance. The seed round, led by Headline with contributions from True, Simon Capital, and returning investor Souschef Ventures, brings its total funding to $10 million. These funds will fuel global expansion and the opening of a San Francisco office. Early deployments show ROI gains, shorter queues, improved upselling, better customer satisfaction, and allow employees to focus on higher-value tasks.
Key Points:
Raised $8.7M in seed funding led by Headline
Total funding now stands at $10M
Supports over 90 languages in drive-thru and mobile ordering
Aims to reduce queues, boost upsells, and assist staff
Takeaway: Vox AI’s funding marks a shift in voice‑first interfaces for quick‑service restaurants, offering scalable, autonomous ordering that eases labor pressures, improves guest experience, and reshapes frontline operations—all without hardware overhauls.
Voicee is auctioning exclusive sector-based rights to its foundational voice AI monetization patents, offering major licensing opportunities.
Voicee has launched a strategic auction offering exclusive licensing rights to its voice AI monetization patent portfolio, branding it as the foundational “Tollbooth” for voice-driven commerce. The portfolio spans sectors like retail, healthcare, AI platforms, AR/VR, and automotive, with market projections exceeding $100 billion by 2030. Bids are structured by sector, with minimums ranging from $35 million to $80 million. Participants must be major players in their respective industries. The auction’s timeline includes packet release in mid-September and final awards in December. Winners gain not just direct implementation, but also sub‑licensing privileges within their sectors.
Key Points:
Patent portfolio covers retail, healthcare, AI, AR/VR, automotive voice commerce
Exclusive sector rights auctioned, not sold outright
Bidding ranges from $35M to $80M per sector, total minimum $255M
Sub‑licensing rights included for winners
Takeaway: Voicee’s auction offers a rare opportunity for leading companies to secure powerful voice AI monetization patents early—potentially shaping control over how voice commerce evolves across key industries.
SoundHound’s generative AI voice assistant now powers natural, context-aware conversations in select Jeep vehicles across Europe.
SoundHound AI has launched its generative AI–driven voice assistant in select Jeep models across European markets. This launches a more intuitive and conversational in-car experience, allowing drivers to engage naturally with their vehicle. The assistant understands context, handles follow-up questions, and elevates what’s possible in in‑cab interaction. Beyond user convenience, this rollout positions SoundHound to tap into a burgeoning voice commerce opportunity in the auto industry.
Key Points:
Live deployment of generative AI voice assistant in Jeep vehicles across Europe
Enables context-aware, multi-turn conversations beyond basic commands
Partnership with Stellantis boosts SoundHound’s automotive AI credibility
Voice commerce integration opens new revenue potential within automakers
Takeaway: This launch shows how generative AI can shift in‑vehicle voice systems from simple command tools into conversational platforms—and ushers in new monetization paths for automakers through voice commerce.
🎙️ Mic Drop
What else is making noise in voice AI.
OpenAI's new speech model intensifies competition, pressuring specialized voice AI startups to innovate faster. (cxtoday.com)
OpenAI highlights expressive, instruction-following voice AI capabilities targeting enterprise adoption and increased developer flexibility. (venturebeat.com)
SuperDial secures $15M funding to scale its healthcare voice AI platform, expanding R&D and commercialization. (mobihealthnews.com)
Loman AI raises $3.5M to enhance voice assistant solutions tailored for the restaurant sector amid increased industry demand. (pulse2.com)
Multiple startups, funding rounds, and product launches highlight surging demand for restaurant-optimized voice AI. (restaurantbusinessonline.com)
NoBroker’s voice-driven AI cloud is being adopted by major Indian enterprises to improve efficiency and customer engagement. (entrepreneur.com)
what3words leverages OpenAI speech models for voice-driven address recognition, aiding logistics and e-commerce operators. (retailtechinnovationhub.com)
KFC’s debut of a multilingual AI order taker in Dubai showcases advances in language support for retail voice systems. (avinteractive.com)
ZONE HSS1 offers a neckband wearable with AI voice interaction, answering questions and translating in real-time, focused on mobility. (thegadgetflow.com)
Google’s free universal translator provides real-time multilingual voice translation, easing global communication challenges. (geeky-gadgets.com)
Step-by-step technical guidance on building voice assistants capable of fluent bilingual interaction. (towardsdatascience.com)
Dreamface launches a next-gen, hyper-realistic voice cloning tool aimed at creators and businesses for video and content production. (ainvest.com)