@FinanceYF5: Sonnet 5 基准测试不如 Opus 4.8，但有一个用法能让它直接起飞：在 Claude Code 中三步设置：/model → Sonnet 5，/effort → Ultracode，复杂任务自动触发 Dynamic Wo…
Summary
Sonnet 5 的基准测试不如 Opus 4.8，但通过 Claude Code 中的三步设置（/model Sonnet 5，/effort Ultracode，复杂任务自动触发 Dynamic Workflow）可以充分发挥其优势，配合 Fable 5 作为智能顾问，实现模型分工协作。
View Cached Full Text
Cached at: 07/02/26, 04:19 AM
Sonnet 5 benchmarks below Opus 4.8, but one usage makes it shine
Three-step setup in Claude Code:
/model → Sonnet 5 /effort → Ultracode Complex tasks automatically trigger Dynamic Workflow
Dan McAteer’s take: Fable 5 is back — now you can use Fable as a super-intelligent advisor and Sonnet 5 as the execution layer, with both models working in tandem. https://t.co/IwNfwpRxHr
Similar Articles
@FinanceYF5: Use Fable 5 as the orchestrator, and use Opus + Codex to execute tasks to save on Fable usage: Fable 5 (highest reasoning strength) = orchestrator, Opus = deep reasoning sub-agent, Sonnet = mechanical execution sub-agent, Codex = peer senior engineer providing different perspectives...
Introduces how to use Fable 5 as the orchestrator, combined with Opus and Codex models to execute tasks to save on Fable usage, including specific configuration in Claude Code.
@FinanceYF5: Claude Code has a powerful new feature: dynamic workflows. How to try: Set /model to opus 4.8, set /effort to “ultracode”, and use “workflow” in your prompt. Claude will write an orchestration…
Claude Code has introduced a dynamic workflow feature that allows users to set parameters and prompts to launch sub-agents, validate results, and report back.
@AYi_AInotes: Just now Boris, the father of Claude Code, shared the most underestimated feature of Claude Code, can directly boost your coding efficiency by 5x! It seems auto mode is truly the killer feature! No permission popups, no more clicking 'allow' every 30 seconds, you can start a session and let it run through the entire...
Boris Cherny, the father of Claude Code, shared the most underestimated feature—auto mode. No permission popups, can run multiple sessions simultaneously, boosting programming efficiency by 5x. Pro plan now supports Sonnet 4.6 and Opus 4.7.
@rohanpaul_ai: Claude Sonnet 5 is more expensive (around +15%) per task than Opus 4.8 and much more expensive (2X) than Sonnet 4.6, ev…
Claude Sonnet 5 costs more per task than previous models due to higher token usage despite lower per-token price, with discounted pricing until August 2026.
Claude Sonnet 5 is out and the gap with Opus 4.8 is smaller than I expected
Anthropic released Claude Sonnet 5, which achieves benchmark scores very close to Opus 4.8 at a significantly lower price, making it a compelling option for agentic tasks despite potential real-world gaps.