Follow ZDNET: Add us as a favorite resource on Google.
Highlights taken by ZDNET
- Opus 5 focuses on coding, knowledge work, and agent workflow.
- Anthropic claims to use several tokens with the same quality.
- Developers get fast output and smart tool controls.
To the surprise of no one, Anthropic released Claude Opus 5. After releasing the controversial Mythos 5 and Fable 5 last month and Sonnet 5 on July 1, it was inevitable that Opus would roll from 4.8 to 5. That is happening today.
I am impressed with Opus 4.8. I’ve been using it for a number of large coding projects, and it handles them quite well. But based on limited user reports, it’s likely that Claude Code will get another big upgrade.
Also: I use Anthropic’s Claude AI tools for very different tasks: How to choose between Models, Code, and Work
I’m not the only one impressed with Opus, based on the praise included in Anthropic’s announcement.
Cristian Rivera, software engineer at Stripe, says, “Opus 5 is a clear production step from Opus 4.8. In one weekend, I gave it the main role of staff over my dev environments: it built its own monitor, dialed each box, and pulled me to get judgment calls.”
Harvey creates AI legal tools. According to Niko Grupen, Harvey’s head of applied research, “Claude Opus 5 is a clear step forward in the work of legal agents compared to previous Opus models, and we have seen great benefits in areas of work such as business administration and arbitration.”
He adds: “We were also impressed with Opus 5’s ability to maintain quality at low imaging levels, achieving similar performance while producing 26% fewer tokens on average compared to Opus 4.8 at high imaging.”
The use of tokens is not just a measure of cost. It is also a measure of how much pressure the AI can handle while working on complex issues. There have been a number of times, even with Opus 4.8’s millionth token context window, when my Code Claude asked for a break and took a nap.
Using several tokens for the same operation means that it may progress over a period of time. That would be helpful.
Next to legend 5 performance
Anthropic says “Opus 5 approaches the power of Claude Fable 5 in many domains at half the price.” The company reports that it has the most robust information processing of any model released, and it brings significant benefits to complex coding. The two will go straight to promoting the work of both Claude Cowork and Claude Code.
Also: Claude Fable 5 is back, but I stick to Opus 4.8 for everyday work: 5 reasons why
The AI development vendor supports testing whether Opus 5 is pulling on Fable 5’s heels in terms of capabilities. Cursor founder Sualeh Asif reports, “Claude Opus 5 brings Fable 5’s intelligence closer to Opus’ speed and cost. In CursorBench, it’s inferior to Fable 5 and has many of the same behaviors. We’re excited to see how developers use it in Cursor.”
Anthropic says it expects the Opus 5 to be a daily driver, especially for business. I’ve been using the heck out of Opus 4.8 on my $100/month Claude Max plan and I’ve never felt a shake. I’ve spent the last week plugging it in between multiple agent runs and I’m still only at 31% of my usage limit for the week, which resets later today.
Most likely, I will move to the Opus 5 as my go-to model and will have more runway to accomplish complex tasks.
Also: I used the Claude Code to unblock a Mac app in 8 hours, but it was more work than magic
While Opus 5 is better for informational work (that kind of table stakes for major version development), it also offers greater independence with less regression, better error detection, and general testing of its work.
What I find most interesting is that Anthropic also mentions how Opus helps with biology and other scientific research. The company says that Opus 5 “is more effective in biological functions compared to Opus 4.8.” Unfortunately, the company hasn’t defined “overclocking” in this context, but we can assume it means better, faster, and cheaper than Opus 4.8.
Anthropic says the price is the same as Opus 4.8. For token API users, it’s $5 for a million input tokens and $25 for a million output tokens. For Claude Max users like me, it will be the default model offered and will have “standard restrictions on all paid plans.”
Also: Claude Fable 5 is back, but I stick to Opus 4.8 for everyday work: 5 reasons why
Anthropic still recommends using the Fable 5 for “your most demanding, day-long freelance projects.” I’m getting ready to add in-app purchases to my iPhone and Mac apps, and I think I might assign that big (and scary) project to Fable to see what it can do for me.
Other announcements
The company also made a series of related announcements today. Here is a quick lightning.
- Quick mod for Opus 5 on Claude’s platform: This is a research preview that enables Opus 5’s full potential for 150% faster token generation at twice the regular price. With Claude Code master user subscriptions, this feature can be purchased for additional usage credits.
- Default failure: Anthropic says, “If the security layer rejects a request, the API easily returns to another model within the same request, so your users get a useful response instead of an error.”
- Changes to the central chat tool: Developers can change which external tools Claude uses during a conversation, without having to cache the information. This allows developers to turn off tools when they are no longer needed, which helps with both cost control and security.
Opus 5 is now available for all tier users who previously had access to Opus 4.8.
Would you use Claude Opus 5 as your daily AI coding model? Let us know in the comments below.
You can follow my daily project updates on social media. Be sure to sign up for my weekly update newsletter, and follow me on Twitter/X at @DavidGewirtzon Facebook at Facebook.com/DavidGewirtz, on Instagram at Instagram.com/DavidGewirtz, on Bluesky at @DavidGewirtz.com, and on YouTube at YouTube.com/DavidGewirtzTV.
