LastAI cover
Sign in
← Daily AI Report한국어로 보기

Tue, Nov 10, 2026

Anthropic releases Claude Fable 5.2 and Mythos 5.2

Claude's new 5.2 models are out, with stronger coding and analysis

  1. A new AI for hard coding and analysis is here
  2. Anthropic has released new models. Anthropic (the AI company that makes Claude) released Claude Fable 5.2 and Claude Mythos 5.2. They reportedly got better at coding, reasoning through problems, and understanding images.
  3. One model, two names. Fable and Mythos are the same model underneath. Fable has strong safeguards against misuse, so it can be opened to all users. Mythos is used by security and life science research groups that apply for access.
  4. Faster and cheaper. It produces answers up to 30% faster than the earlier 5.1. The price is also reportedly lower than before. It can reportedly check its own work and keep going on hard tasks for hours.
  5. What it did during testing. It helped finish a software migration that engineers had found hard to solve. It reportedly designed three electronic devices from scratch. It reportedly solved two open problems in black hole physics.
  6. Both excitement and doubt. Matt Wolfe (who covers AI news on YouTube) shared the news quickly on release day. One developer said that after trying it, the models' levels felt evenly matched. Others suspected that older models' performance was deliberately lowered. Some comments questioned whether the comparison charts were fair.
  7. How work may change. Handing long coding jobs to AI may get easier. Security or life science questions may be refused because of the safeguards. In that case, you can ask again with another model.
  8. List of sources
Read as text

What happened?

Anthropic has released new models

  • Anthropic (the AI company that makes Claude) released Claude Fable 5.2 and Claude Mythos 5.2.
  • They reportedly got better at coding, reasoning through problems, and understanding images.

Good to know first

One model, two names

  • Fable and Mythos are the same model underneath.
  • Fable has strong safeguards against misuse, so it can be opened to all users.
  • Mythos is used by security and life science research groups that apply for access.

What's new?

Faster and cheaper

  • It produces answers up to 30% faster than the earlier 5.1.
  • The price is also reportedly lower than before.
  • It can reportedly check its own work and keep going on hard tasks for hours.

What's new?

What it did during testing

  • It helped finish a software migration that engineers had found hard to solve.
  • It reportedly designed three electronic devices from scratch.
  • It reportedly solved two open problems in black hole physics.

How people reacted

Both excitement and doubt

  • Matt Wolfe (who covers AI news on YouTube) shared the news quickly on release day.
  • One developer said that after trying it, the models' levels felt evenly matched.
  • Others suspected that older models' performance was deliberately lowered.
  • Some comments questioned whether the comparison charts were fair.

What changes?

How work may change

  • Handing long coding jobs to AI may get easier.
  • Security or life science questions may be refused because of the safeguards.
  • In that case, you can ask again with another model.

Sources

AI wrote this report from public sources and checked it against them. Please see the original articles for exact details.