AI Trends

Trending story Continuously updated

Yuchen Jin says GPT-6.2 Astra leads by 5.5 points on Agents' Last Exam 1.1

2 reports2 story updatesUpdated Nov 18 02:16

Understand this story

AI overview

On November 18, 2026, Yuchen Jin offered the assessment that GPT-6.2 Astra scored 5.5 points higher than Claude Fable 5.2 on Agents' Last Exam 1.1, at only 5% of the latter's cost. He therefore argued that intelligence-per-dollar should be increased, and the tweet included a link to an image. On the same day, a report by Vaibhav Srivastav stated that GPT-6.2 Astra set a new record on Agents' Last Exam, which tests multi-step professional work, scoring 5.2 points higher than GPT-6 Astra at less than one-third of its cost, and 5.5 points higher than Claude Fable 5.2 at only 5% of its cost. The report said the results were cited from a tweet introducing GPT-6.2 Astra and Sol, and that both models have launched on ChatGPT Work, Codex, and the OpenAI API. There have been no further developments so far.

AI-generated from public reports

Latest development · Report

GPT-6.2 Astra Sets a New Record on Agents’ Last Exam

Developments

1 confirmed developments

  1. Yuchen Jin says GPT-6.2 Astra leads by 5.5 points on Agents' Last Exam 1.1

    Source:Vaibhav Srivastav

Reports and discussion timeline

Browse reports and discussion by time. Their evidence levels remain distinct.

Nov 18

  1. Vaibhav SrivastavSelected

    GPT-6.2 Astra Sets a New Record on Agents’ Last Exam

    GPT-6.2 Astra set a new record on Agents’ Last Exam, which tests multi-step professional work, scoring 5.2 points higher than GPT-6 Astra at less than one-third of its cost; it also scored 5.5 points higher than Claude Fable 5.2 at only 5% of its cost. The results come from a tweet introducing GPT-6.2 Astra and Sol, both of which are now available on ChatGPT Work, Codex, and the OpenAI API.

  2. Yuchen Jin

    GPT-6.2 Astra leads Claude on Agents' Last Exam 1.1…

    Yuchen Jin says GPT-6.2 Astra scores 5.5 points higher than Claude Fable 5.2 on Agents' Last Exam 1.1, at only 5% of the latter's cost. From this he argues that intelligence-per-dollar should be improved; the tweet includes a link to an image.

Heat trend for this story

Current
10
Period peak
20 Nov 18 03:00
24-hour change
—

Move the pointer or select the chart to inspect hourly heat. Use the left and right arrow keys with a keyboard.

Related stories

  1. Related storiesQwen Releases Qwen4-Max Preview Version
  2. Related storiesAnthropic releases Claude Fable 5.2 and Mythos 5.2
  3. Related storiesOpenAI releases GPT-6.2 Astra and Sol, ARC Prize announces perfect scores
  4. Related storiesAnthropic launches Claude Fable 5.2 and Claude Mythos 5.2