← All models

GPT-Live-1.1

Published Publisher OpenAI

Overview

GPT-Live-1.1 is a full-duplex voice model and the successor to GPT-Live-1. OpenAI announced its rollout in ChatGPT Voice on October 15, 2026. It is the default voice model for Plus and higher-tier users, including in Codex and Dots. GPT-Live-1.1 Quip is a separate, smaller variant. (help.openai.com)

Capabilities and uses

  • Listens and speaks continuously, handling pauses, interruptions and changes of pace without requiring a clearly defined turn to end.
  • Supports natural spoken conversations, questions and brainstorming. In ChatGPT Live, available features can include web search, memory, text and images, visual results, plugins and connected apps, depending on the account.
  • The release introduced more natural accents across languages and additional voice options. Voice availability varies by region and account. (deploymentsafety.openai.com)

Limitations

  • ChatGPT Live does not support video or screen sharing. It is primarily designed for one-on-one conversation; background noise, overlapping speech and network conditions can affect what it hears. Transcripts may not be verbatim.
  • Responses can be inaccurate. OpenAI reported statistically significant regressions versus GPT-Live-1 in some safety evaluation categories, including sexual content, mental health, emotional reliance and gore. The model itself did not have code-execution capability at launch; delegated models and tools have separate capabilities and safeguards. (help.openai.com)

Pricing and access

API pricing: Unverified for GPT-Live-1.1 specifically. As checked on November 15, 2026, the official API pricing page lists GPT-Live-1 voice-session pricing, but does not establish that the same rate applies to GPT-Live-1.1. Separate input and output API rates for GPT-Live-1.1 could not be verified. (developers.openai.com)

ChatGPT Voice in Chat: OpenAI lists Plus access at 3 hours per rolling 24-hour period; Pro at USD $100/month includes 15 hours, while Pro at USD $200/month lists unlimited use subject to abuse guardrails. For Enterprise on usage-based USD pricing, Voice in Chat is USD $0.05 per minute; this is a session-duration price, not separate input/output token pricing. Other plans and interfaces have different limits or billing. Checked November 15, 2026. (help.openai.com)

Official sources