Drew Breunig discusses how the release of the Fable model shifted coding strategies by making cost a significant factor, ending the era of easily improved coding harnesses due to model advancements.
# A quote from Drew Breunig
Source: [https://simonwillison.net/2026/Aug/23/drew-breunig/](https://simonwillison.net/2026/Aug/23/drew-breunig/)
23rd August 2026
> Prior to Fable, it felt silly to waste*too*much time improving your coding harness or context strategies\. A new model would arrive at the same price \(or cheaper\!\) and paper over most of your problems\. But then Fable landed\. It was \(and still is\!\)*incredible*\. But the cost was so high and Opus was*good enough*\(as was 5\.6, K3, and even GLM\) for*most*of the code we needed\. *So we started to think about what work went where\.*
—[Drew Breunig](https://www.dbreunig.com/2026/08/23/fable-the-end-of-moore-s-law.html),Fable & The End of the Free Lunch
Posted[23rd August 2026](https://simonwillison.net/2026/Aug/23/)at 7:55 pm
## Recent articles
- [Conceptual integrity and counting lines of code](https://simonwillison.net/2026/Aug/19/conceptual-integrity-and-counting-lines-of-code/)\- 19th August 2026
- [Qwen 3\.8 27B is excellent, but it defaults to wildly overthinking things](https://simonwillison.net/2026/Aug/16/qwen-38-27b/)\- 16th August 2026
- [Now we have a timeline of the OpenAI accidental attack against Hugging Face](https://simonwillison.net/2026/Aug/7/openai-timeline/)\- 7th August 2026
This is a**quotation**collected by Simon Willison, posted on[23rd August 2026](https://simonwillison.net/2026/Aug/23/)\.
[ai2,200](https://simonwillison.net/tags/ai/)[generative\-ai1,949](https://simonwillison.net/tags/generative-ai/)[llms1,916](https://simonwillison.net/tags/llms/)[anthropic329](https://simonwillison.net/tags/anthropic/)[claude301](https://simonwillison.net/tags/claude/)[drew\-breunig21](https://simonwillison.net/tags/drew-breunig/)[llm\-pricing90](https://simonwillison.net/tags/llm-pricing/)[claude\-mythos\-fable37](https://simonwillison.net/tags/claude-mythos-fable/)
The article reflects on how the release of high-cost AI models like Fable is ending the 'free lunch' era in software development, forcing developers to optimize by strategically selecting between models based on cost and performance, similar to the impact of Moore's Law slowdown.
Cognition replaced the Opus model with Fable in Devin's Fusion architecture, achieving higher performance at lower cost despite Fable's higher per-token price, through better delegation and reduced lead model turns.
Anthropic benchmark shows that using a larger model (Fable) as orchestrator with cheaper models (Sonnet) as workers achieves 96% of full Fable performance at 46% cost, available now in Claude Code.
Anthropic released Fable 5, a powerful new model with high pricing, making cost-aware routing essential for agent builders due to token fan-out and high output costs.
A practical tip from Simon Willison: allow Claude Code's Fable model to use its own judgment for testing and delegate coding tasks to subagents running cheaper models, reducing token usage while maintaining quality.