LLM Digest
Subscribe

AI Storyline

7 items · 6 sources · 6 days

View as JSON

Operational story trace

Opus 5.5

Current stateShippingstatus changed Sep 29

Latest change

Latent Space reports Opus 5.5 handles explainer-video generation well, an unusually concrete capability read that follows Anthropic's Sonnet 5.5 release on AWS on Sep 28.

Earlier contextThe story so far

Anthropic released Claude Opus 5.5 on Sep 22, with same-day availability on Amazon Bedrock, and OpenAI answered about an hour later with GPT-6 Sol and Luna. Pricing cuts of 40-50% across vendors followed, and Anthropic then shifted its pitch to cost per task rather than token price.

editor-curated · source-linked

State over time

● launch + price war · Sep 22–23Sonnet 5.5 + hands-on · Sep 28 → now ●
  • launch + price war · Sep 22–23
  • cost-per-task framing · Sep 24–25
  • Sonnet 5.5 + hands-on · Sep 28 → now
LAUNCH · Sep 22
Opus 5.5 ships on Bedrock the day Anthropic announces it, minutes before GPT-6 Sol and Luna
2 sources · show sources ▾
PRICE WAR · Sep 23
Vendors cut prices 40-50% and Opus 5.5 becomes AINews's default model
1 source · show source ▾
THE TURN · Sep 24–25
Anthropic argues for cost per task over token price
Two Claude blog posts tie Opus 5.5 pricing to longer, context-heavy coding sessions and show which Claude Code settings change the bill.
2 sources · show sources ▾
NOW · Sep 28–29
Sonnet 5.5 lands on AWS and Opus 5.5 gets its first niche capability read
2 sources · show sources ▾

What to watch — open questions

  • Do independent benchmarks confirm Anthropic's lower cost per task for Opus 5.5 versus GPT-6 Sol?
  • Do the 40-50% price cuts hold, or are they tied to launch promotions?
  • Does Sonnet 5.5 undercut Opus 5.5 on cost per task for coding workloads?
How this thread was built
editor wrote the arc · 4 beatswatcher 1 status change

Storylines are threaded mechanically from the feed: stories that share a distinctive anchor across multiple days and sources. Each item links to its original source. The evidence trace, current state, and open questions are written by the editor routine and refreshed whenever a new beat lands.