Overview
GPT-Live-1 mini is a voice model in OpenAI’s GPT-Live family, designed for natural, real-time conversations. At launch, OpenAI designated it as the default ChatGPT Voice model for Free users. (deploymentsafety.openai.com)
Capabilities and uses
- Uses full-duplex interaction to listen and speak continuously, handle pauses and interruptions, and adjust conversational timing.
- Can delegate more complex reasoning or search to a backend model while continuing the conversation. At launch, the mini version used GPT-5.5 Instant in the background; OpenAI stated that backend models could change.
- Intended uses include hands-free assistance, language practice, storytelling, conversation and live translation. These are GPT-Live family and ChatGPT Voice use cases; the announcement does not provide a separate feature matrix for mini. (openai.com)
Limitations
- At launch, GPT-Live in ChatGPT did not support voice with video or screen sharing. Some languages could have non-native accents or fluency gaps.
- ChatGPT uses predefined voices and includes safeguards against impersonating real people.
- At launch, the voice models lacked broad independent tool access and code execution; complex work could instead be delegated to other models. (openai.com)
Release and availability
Announced on July 8, 2026, with a gradual rollout to ChatGPT users. The launch announcement said API access was planned; it did not establish a mini-specific API identifier or availability. (openai.com)
Pricing
As of November 16, 2026, a separate official API price for GPT-Live-1 mini could not be verified. Its input and output billing rates, currency and billing units are therefore unverified. OpenAI’s API pricing page lists a session price for GPT-Live-1, but that is a different model and is not attributed to the mini version here. (developers.openai.com)