Rafomac News, Tech & Trends That MatterClaude Opus 5 has dropped, and it’s shaking up expectations by outperforming Fable 5 on almost every benchmark—while charging half the price. But the win comes with some big caveats that might change how you think about using it.
Why Claude Opus 5 Feels Like a Game-Changer
Claide Opus 5 didn’t just arrive quietly—it made a bold statement by beating Claude Fable 5 in nearly every aspect. It’s faster, cheaper, and offers full access to your allotted usage, unlike Fable 5, which frustratingly restricts you to only halfway through your limitations.
This price advantage is huge. Opus 5 costs roughly half as much as Fable 5, making it far more accessible to the average user. Plus, it runs lightning-fast. These are not minor tweaks; they are leaps forward that force you to ask: is there any reason left to use Fable 5?
The New Finn Benchmarks Put Claude Opus 5 in the Spotlight
To prove its edge, a rigorous set of tests — the Finn benchmarks — were run comparing Fable 5 and Opus 5 across five demanding tasks. From building a detailed 3D roller coaster simulation to cloning the Apple website pixel-by-pixel, Opus 5 consistently demonstrated its superior capabilities.
Take the roller coaster test: Opus 5 generated a beautiful, dynamic simulation with rich detail including trees, sky, and real-time ride perspective, while Fable 5’s output felt dated and less impressive. And though Fable 5 used fewer tokens in some cases, its cost was about 50% higher.
What Opus 5 Does Better—and Where It Still Falters
In a demanding agentic test that required sifting through various documents and files, Opus 5 scored 5 out of 8 correct scavenger hunt answers, while Fable 5 ran into content-usage blocks and couldn’t complete the task at all. For bug detection and fixing in open-source code, Opus 5 took slightly more time but remained about 25% cheaper overall.
The only test that slightly favoured Fable 5 was the bridge-building challenge, where it engineered a structure to hold marginally more weight than Opus 5’s design—though at nearly double the cost.
Three Deal-Breakers That Temper the Hype
Despite these wins, Opus 5 comes with three major drawbacks that can’t be ignored. First, its personality has been described as nearly unbearable. Instead of the easy, engaging conversations Claude models are known for, Opus 5’s responses are overly verbose, unfocused, and sometimes feel like a know-it-all friend who won’t stop talking. This verbosity can even interfere with simple tasks like bug fixing, where the model digresses instead of acting efficiently.
Users have found themselves repeatedly needing to stop its output and simplify instructions, even resorting to editing Opus 5’s personality prompts to speak “as simple as humanly possible.” It’s a jarring shift that might discourage anyone seeking a smooth interaction.
Second, the usage limits remain a sore spot. Although Opus 5 lets you use 100% of your budget, the total available usage is still much lower than ChatGPT, where limits can be refreshed every few minutes. This restriction injects an anxiety about prompt economy that dampens creativity and exploration.
Finally, the supporting ecosystem—the so-called harness around the model—is still behind what ChatGPT offers. The newly introduced Claude voice mode is a nice addition but nowhere near the polished experience of ChatGPT’s voice features.
Where Opus 5 Fits in a Smart AI Stack
Putting it all together, Opus 5 shines in raw intelligence and cost efficiency, making it a top choice for tackling complex, hard problems. Yet for planning, brainstorming, or any back-and-forth conversation where personality matters, many still prefer Fable 5 despite its higher costs and usage restrictions.
The overall daily driver, at least for now, remains ChatGPT 5.6, thanks to higher usage limits and a more robust voice mode.
The launch of Opus 5 promises exciting new possibilities—but its quirks remind us that in AI, power isn’t just about intelligence or speed; usability counts just as much.
