←── back to feed
/topics/claude-opus-5-release-and-capabilities

Claude Opus 5 release and capabilities

42 items3 sourcesupdated 8d agotrend 0

Anthropic released Claude Opus 5 on July 24, 2026, a hybrid reasoning model offering improved performance on coding and knowledge tasks at the same cost as Opus 4.8. The model achieves near-Fable 5 capabilities at half the price and is now the default on Claude Max and strongest model on Claude Pro.

  • Released July 24, 2026; default model on Claude Max and strongest on Claude Pro
  • 200K token context window by default; improved performance on coding and knowledge tasks
  • Achieves Fable 5-level frontier intelligence at half Fable 5's price, matching Opus 4.8 cost
  • Claude Code system prompt reduced by over 80% for Opus 5; hardcoded instruction prevents subagent use
  • Extended thinking no longer shown to users by default
[HN]hacker news23
Opus 5: RL-Fried and mistake-prone for anyone else?
HN: Claude · ValentineC · ▲5 · 9d
Claude Fable 5: What It Is and Why the Hype?
HN: Claude · lazerg · ▲1 · 9d
Benchmarking Opus 5 on SlopCodeBench
HN: GPT · dhorthy · ▲198 · 10d
Claude Opus 5 is limited to a 0.2M token context window by default
HN: Claude · colejohnson66 · ▲2 · 10d
Elevated Errors for Opus 5
HN: Claude · TimCTRL · ▲130 · 11d
Claude Code has a hardcoded instruction telling Opus 5 not to use subagents
HN: Claude · bredren · ▲33 · 11d
Claude no longer shows full thinking
HN: Claude · rzk · ▲5 · 12d
Claude Opus 5: The System Card
HN: Claude · paulpauper · ▲4 · 12d
I was paying too much for Claude and codex here is how I reduced 90% of it
HN: Claude · Cognifyr · ▲3 · 12d
The Many Claudes of My Code
HN: Claude · u8 · ▲3 · 12d
Claude Code Lightbar for Mac OS
HN: Claude · gw5815 · ▲4 · 12d
Show HN: Curated Claude Code – a small agent harness with an intake gate
HN: Claude · Aleksandr_NFA · ▲4 · 12d
Show HN: Impressive turn-based RPG prototype built with Opus 5 in 20 min
HN: Claude · sherujgr · ▲4 · 12d
MCP Clock: live public hosted tool for Claude, Codex, any MCP Client
HN: Claude · firasd · ▲3 · 12d
Show HN: AI Toolbox supports Claude Opus 5
HN: Claude · superduper5551 · ▲2 · 12d
Show HN: Sync Claude Code and Codex configs, with a board that shows the drift
HN: Claude · slash9494 · ▲2 · 12d
"We removed over 80% of Claude Code's system prompt for Opus 5 and Fable 5"
HN: Claude · nreece · ▲5 · 12d
AIs don't do what you want. This is bad
HN: AI · kking23 · ▲99 · 13d
Opus 5 System Prompt
HN: GPT · miohtama · ▲4 · 13d
Claude Opus 5 pricing and launch-day cost-per-task result
HN: Claude · alexmercerdev · ▲4 · 13d
The new rules of context engineering for Claude 5 generation models
HN: Claude · opwizardx · ▲10 · 13d
What's new in Claude Opus 5
HN: Claude · acmnrs · ▲7 · 13d
Claude Opus 5
HN: Claude · alvis · ▲1.8k · 13d
[BSKY]bluesky8
I've already had to update the guide to which AI models to use that I wrote on Thursday to include Opus 5 and Codex's voice mode, both of which are significant & launched on Friday. Keeping up is challenging, even if you are following this…
@emollick · @emollick.bsky.social · ▲76 · 11d
In my testing, Opus 5 replaced Opus 4.8 for me, it was generally stronger in everything... except it shares some of the weird language quirks of Fable, including a love of density and FableSpeak.
@emollick · @emollick.bsky.social · ▲54 · 13d
I had access to Opus 5 before release and found it to be a good model if a quirky one. On shorter tasks, it could match or beat Fable levels of performance, at longer tasks it seemed less ambitious & would not deliver as complete a set of …
@emollick · @emollick.bsky.social · ▲103 · 13d
Insane numbers for opus 5, the power of faster iteration speed + scaled RL (Fable too big to RL as well, yet). And on safeguards "Based on our testing, we expect the classifiers to intervene around 85% less often than they do for Fable 5.".
@natolambert · @natolambert.bsky.social · ▲89 · 13d
Unexpected finding from a study on ChatGPT's impact on college: "once the COVID-19 disruption is modeled separately, the introduction of ChatGPT had no detectable effect on grades.. course evaluations for subject understanding, interest, a…
@emollick · @emollick.bsky.social · ▲67 · 13d
The real comparison to Moore's law for AI isn't scaling laws, but rather the intelligence efficiency that we gain year-over-year using models to get better at training models.
@natolambert · @natolambert.bsky.social · ▲22 · 13d
As a joke I prompted Codex "Build and run BenchBench, a benchmark of now good ai is at creating benchmarks. then figure out what benchbenchbench is and run that. and then write benchbenchbench up as a good arXiv paper." I got a PDF.
@emollick · @emollick.bsky.social · ▲87 · 13d
Glad to see Google sharing data on how Gemini is being used. Especially interesting is that the usefulness of multimodal AI for manual labor may be greater than expected. blog.google/innovation-a...
@emollick · @emollick.bsky.social · ▲79 · 13d