I keep seeing people stress about not having access to the newest, shiniest AI model. So when an AI professional posted a clear-headed take on this, I had to share it. The short version? You don’t need Fable 5 to do great work. The other Claude models are more than enough.
What I love about this breakdown is how practical it is. The author isn’t chasing hype. They’re comparing tools head to head and telling you exactly where each one wins.
Fable 5 vs Opus 4.8: the real tradeoff
Here’s the comparison that stopped me in my tracks. In the creator’s own tests, Fable 5 costs twice the tokens of Opus 4.8 for only slightly better output on creative use cases.
Think about that. You’re paying double for a small bump. For most work, that math doesn’t add up. The expert’s point is simple: match the model to the job and you’re set.
- Haiku: the quick, lightweight tasks
- Sonnet: everyday work
- Opus: the hard, complex problems
No single model needs to do everything. Use the cheap, fast one when speed matters. Save the heavy hitter for the genuinely tough stuff. That’s how you keep quality high and costs sane.
Claude as the operating system, with backups ready
This is where the post gets really useful. The original poster treats Claude as their operating system, but keeps alternatives ready for every part of it. Smart, because no tool should be a single point of failure.
Here’s the side-by-side they shared:
- Claude Chat for slides, docs, research, prompts and story design, with ChatGPT or Gemini as backups
- Claude Code for file work and agentic workflows on your computer, with Codex as the alternative
- Claude Design for turning raw data into websites, slides and brochures, with Google AI Studio or Manus standing by
- Claude Connectors for images and videos via MCP, with ChatGPT or Codex covering the gap
- Claude Skills for packaging reusable workflows, still leading the pack, though the creator notes others are catching up
I think this is the part most people miss. Picking a main tool isn’t about loyalty. It’s about knowing your options so you’re never stuck.
Why the human part matters more, not less
The mind behind this post made an observation that stuck with me. AI used to give advice while you did the execution. Now Claude does the execution, and you sit in the director’s seat.
So the human part matters more, not less. The context. The thinking. The judgment.
That flips the usual fear on its head. The tool getting better doesn’t make you less important. It raises the bar on the things only you can bring: taste, direction, and good calls.
A quick look at the workflow in action
To show it working, the contributor shared a real example. They combined their website as a style reference, a voice note as the content reference, and Runway MCP for images and videos, then fed it all into Claude Design for the final output.
That’s a neat little stack. Style in, content in, media in, polished result out. You can copy the pattern for your own projects today.
My takeaway
If I had to sum up the recommendation, it’s this: stop chasing the priciest model and start matching tools to tasks. Use Haiku, Sonnet, and Opus for what each does best. Keep a backup ready for every workflow. And remember the judgment is still yours.
The full post has more nuance and the creator’s exact example setup. Check out the original LinkedIn post for the complete breakdown.