Model

gpt-5.2

6 entries · December 2025 – July 2026

GPT-5.2 was an update to the GPT-5 line released in December 2025, pitched around professional knowledge work rather than one headline capability; OpenAI reported it tying or beating human professionals on 70.9% of comparisons on its internal GDPval benchmark, up from 38.8% for GPT-5.1. It arrived roughly three weeks after Google's Gemini 3 took a lead on independent leaderboards and, according to later reporting, prompted an internal "code red" at OpenAI. As with the rest of the series, its benchmark figures were self-reported rather than independently verified.

Appears alongside

Featured in threads

Tracks

  • Models & capabilities 4
  • Benchmarks & progress 2
  • Open weights & ecosystem 1

  1. US CAISI publishes assessment of Z.ai's GLM-5.2

    The US assessment found GLM-5.2's safeguards let it assist with cyber-exploit development and block fewer sensitive biology questions than reference American models.

    Benchmarks & progress · Open weights & ecosystem

  2. OpenAI retires GPT-4o and other older models from ChatGPT

    OpenAI said only about 0.1% of ChatGPT users still chose GPT-4o daily, more than a year after briefly restoring it to Plus users following the GPT-5 rollout backlash.

    Models & capabilities

  3. OpenAI launches Prism

    Built on Crixet, a LaTeX platform OpenAI had quietly acquired, Prism is free for any ChatGPT account and handles citation management and sketch-to-LaTeX conversion.

    Models & capabilities

  4. OpenAI launches OpenAI for Healthcare

    The HIPAA-compliant suite, built on GPT-5.2, launched with initial deployments at AdventHealth, Cedars-Sinai, HCA Healthcare, Memorial Sloan Kettering, Stanford Medicine Children's Health and UCSF.

    Models & capabilities

  5. OpenAI ships GPT-5.2-Codex

    OpenAI reported an 'unmatched' 56.4% on the SWE-Bench Pro benchmark and 64% on Terminal-Bench 2.0, alongside new defensive-cybersecurity capabilities.

    Models & capabilities

  6. OpenAI introduces FrontierScience benchmark

    GPT-5.2 scored 77% on olympiad-style questions but 25% on open-ended research tasks, a gap OpenAI's own researchers said showed little improvement over GPT-5.

    Benchmarks & progress