--- title: "FABLE IS BACK! (And Sonnet 5 is here too)" url: https://daily.dev/posts/fable-is-back-and-sonnet-5-is-here-too--n869tjaw2 source_url: https://www.youtube.com/watch?v=KSV-7ywHxeU type: video:youtube source: "Theo - t3․gg" published: 2026-07-01T04:07:00.514Z updated: 2026-07-01T22:11:19.698Z tags: ["ai-agents", "claude", "anthropic"] reading_time: 28 upvotes: 0 comments: 0 language: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # FABLE IS BACK! (And Sonnet 5 is here too) **[Theo - t3․gg](https://daily.dev/sources/t3dotgg)** · 28 min read · 0 upvotes · 0 comments ## Summary A hands-on evaluation of Claude Sonnet 5, covering its benchmarks, real-world performance, and cost efficiency compared to Opus 4.8, GPT-5.5, and other models. Key findings: Sonnet 5 introduces Fable-like sub-agent orchestration capabilities to a mid-tier model, but it is inefficient — using far more tokens than competitors, running slowly, and often costing more than Opus in practice despite lower per-token pricing. The model's agentic behaviors (spawning sub-agents, breaking up tasks) are promising but undermined by insufficient intelligence, causing it to loop and waste tokens. Also covered: the lifting of US export controls on Claude Fable 5 and Mythos 5, benchmark contamination concerns (especially SWE-bench), and a leaked thinking trace showing verbose, unfocused internal reasoning. The author concludes Sonnet 5 is best suited as a sub-agent called by smarter orchestrators, not as a daily coding model. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#claude](https://daily.dev/tags/claude), [#anthropic](https://daily.dev/tags/anthropic) [View this post on daily.dev](https://daily.dev/posts/fable-is-back-and-sonnet-5-is-here-too--n869tjaw2)