Opus 5.5 uses 7% of weekly quota in 1h40m run, user compares it with GPT-6 Sol/Luna
The author ran Opus 5.5 at max effort for 1 hour 40 minutes and 2.3 million tokens at 7% of the weekly quota, and says GPT-6 Sol/Luna chase cheap intelligence while Chinese open-source model Mimo steals attention.
Even the guys at Anthropic admit it: Claude Opus 5 was terrible. But (oh my god) Opus 5.5 is a beast: I used Opus 5.5 at max effort for 1 hour 40 minutes. It used 2.3 million tokens, but it's still running. And my weekly usage is at 7%. It used to take 50%. My two cents on Opus 5.5 & GPT-6 Sol/Luna: 1. Opus 5.5 is better than Fable 5.1. Way cheaper too. And the higher rate limits make a massive difference when you actually use it for hours. 2. For my work, I don't see a reason to use another Claude model. Just switch between effort. Only use Opus 5.5. No more Fable (until Fable 5.5?). 3. Paying more doesn't always get you better AI. I know it feels like Fable is smarter. My bet: Fable 5.5 (or 6) already exists. They haven't released it yet. The next Fable will give us a reason to switch back. 4. ChatGPT goes after super cheap intelligence with Sol/Luna 6. You don't need the smartest AI for every task. For everyday work, GPT-6 Sol/Luna is good. But, but, a new Chinese open-source model ruined the party for ChatGPT. It's called Mimo: 10x cheaper than Sol-6 and much better than Luna-6. It's a battle. And we are getting new models every Tuesday. I'm tired, boss. To not miss out on the next release (and by missing out, I mean HOW TO USE IT not the news itself), I write the newsletter "How to AI". 955,000 readers. Free: http://how-to-ai.guide.
This post quotes another post: x.com/rubenhassid/status/2082708188005298461