OpenAI's GPT-Live-1 API lets developers build apps that talk and listen at the same time
decoder ·
OpenAI has made its GPT-Live-1 speech model available to developers through an API. The model supports full-duplex interaction, meaning it can listen and speak simultaneously, and it is already used inside ChatGPT. Developers can combine it with different backend models to balance reasoning depth, speed, and cost for each use case. The API is priced at $0.05 per minute, which the article describes as not cheap. Yelp is using the model for phone-based reservations and reports improved call handling, according to CTO Alex Levy. On OpenAI's benchmarks, GPT-Live-1 scores 80.1 percent in full-duplex interactivity tests, compared with 45.4 percent for GPT-Realtime-2.1, while turn-taking latency falls to 0.8 seconds from 1.4 seconds. Tool-calling accuracy rises to 87 percent from 60 percent, and a banking voice support benchmark shows a 32 percent pass rate versus 12.4 percent for the previous model. GPT-Live-1 also includes twelve new voices covering different accents, dialects, and languages, and it provides ASR transcripts and response text out of the box, with full details promised in the API documentation.
decoder