CyberHappenings logo
☰
🏠 Home 📂 Browse ℹ️ About

Track cybersecurity events as they unfold. Sourced timelines, daily updates. Fast, privacy‑respecting. No ads, no tracking.

OpenAI Introduces Thinking Effort Picker for ChatGPT

First reported
Last updated
📰 1 unique sources, 1 articles

Summary

Hide ▲

OpenAI is testing a new feature called the Thinking effort picker for ChatGPT. This feature allows users to select the intensity of the model's thinking process. The picker has four levels: light, standard, extended, and max. The max level is only available to Pro plan subscribers. The feature aims to give users more control over the model's responses based on the complexity of the queries. The levels are designed to balance response depth and speed, with higher levels providing more detailed answers but taking longer to respond. The picker is currently being tested and is not yet widely available.

Timeline

  1. 31.08.2025 14:26 📰 1 articles · ⏱ 16d ago

    OpenAI tests Thinking effort picker for ChatGPT

    OpenAI is introducing a new feature called the Thinking effort picker for ChatGPT. This feature allows users to select the intensity of the model's thinking process, with four levels available: light, standard, extended, and max. The max level is restricted to Pro plan subscribers. The feature is currently in testing and aims to provide users with more control over the model's responses based on query complexity.

    Show sources
    Open in new tab

Information Snippets

  • The Thinking effort picker allows users to choose the intensity of ChatGPT's thinking process.

    First reported: 31.08.2025 14:26
    📰 1 source, 1 article
    Show sources

  • There are four levels of thinking effort: light, standard, extended, and max.

    First reported: 31.08.2025 14:26
    📰 1 source, 1 article
    Show sources

  • The max thinking level is available only to Pro plan subscribers.

    First reported: 31.08.2025 14:26
    📰 1 source, 1 article
    Show sources

  • Higher thinking levels provide more detailed responses but take longer to generate.

    First reported: 31.08.2025 14:26
    📰 1 source, 1 article
    Show sources

  • The feature is currently in testing and not yet widely available.

    First reported: 31.08.2025 14:26
    📰 1 source, 1 article
    Show sources