How Anthropic Engineers Actually Prompt Fable 5
A guide to effectively prompting Claude Fable 5, covering six key techniques including providing context/intent, using negative prompting, letting the model act once ready, verifying outputs, avoiding reasoning requests, and keeping instructions concise. The speaker emphasizes that Fable 5 is expensive and should only be used 5-15% of the time, with pricing at $10 per million input tokens and $50 per million output tokens.
Summary
The transcript features a detailed guide on how to efficiently use Claude Fable 5, Anthropic's latest model. The speaker begins by establishing that Fable 5 is the strongest model they've used but emphasizes it's a premium-priced model costing double Opus. During a promotional period lasting only until July 7th, users get 50% of weekly limits at no extra cost, after which they must switch to paid usage credits. The model is accessible across multiple platforms including Claude's web interface, desktop app, VS Code, and other integrations.
The core of the guide presents six key prompting habits. First, providing context and intent behind requests allows Fable 5 to understand the broader task and connect to relevant information rather than guessing. Second, using negative prompting—explicitly telling the model what not to do—prevents the model from over-reaching with creative but unwanted additions, similar to how you'd instruct an intern on specific boundaries. Third, the speaker advises stopping overplanning and letting the model act once it has sufficient information, rather than requiring extensive pre-planning.
The fourth technique is verification loops: making the model prove its work before claiming completion. The speaker stresses that trust is built through multiple verification cycles, not blind acceptance of outputs. Fifth, and specific to Fable 5, the speaker warns against asking the model to show its reasoning, as this can trigger safety guardrails that silently route requests to the less capable Opus 4.8 model. Finally, the speaker advocates for conciseness, arguing that Fable 5's improved intelligence means short, clear instructions can be as effective as lengthy, rule-filled prompts.
The speaker also addresses Fable 5's safety architecture, explaining that built-in safeguards will automatically route requests related to hacking, dangerous biology, or attempts to expose internal reasoning to Opus 4.8 without user notification (though API users will see which model responded). The speaker concludes by recommending users read Anthropic's official documentation and emphasizing that Fable 5 should be reserved for only the most complex tasks (5-15% of use cases) to avoid unnecessary costs.
Key Insights
- The speaker claims that Fable 5 is the strongest model they've ever used, built differently from Opus and GPT, and follows short, clear directions better due to improved reasoning capabilities that make it feel like it understands intent more deeply.
- According to the speaker's analysis, Fable 5 costs double Opus at $10 per million input tokens and $50 per million output tokens, and should realistically only be used 5-15% of the time to avoid excessive costs, making it overkill for routine tasks.
- The speaker reports that asking Fable 5 to show its reasoning or explain itself in system prompts can trigger safety guardrails that silently route the request to the less capable Opus 4.8 model, representing a Fable-specific prompting constraint.
- The speaker claims that Fable 5 has built-in safety guardrails that automatically route requests related to hacking, dangerous biology, or attempts to expose internal reasoning to Opus 4.8 without user notification, though API users will see which model responded.
- The speaker argues that despite conventional wisdom, shorter instructions can now steer Fable 5 as effectively as verbose, rule-laden prompts because the model's improved intelligence allows it to infer intent from minimal guidance.
Topics
Transcript
[0:00] So, we've all waited long enough. Fable 5 is finally coming back to us, and it's an incredible model. It's the strongest one that I've ever used, hands down. I've built things like my second brain, my AI operating system. So, obviously, I've been playing around with this model a ton, and I've been trying to figure out the best way to use it so that you can actually use it efficiently and you're not paying for tokens for no reason. I've looked at what people have said on X. I've listened to Enthropic Engineers, and also, I've read through this entire documentation right here on the best practices for prompting Claude Fable 5. So today what I did…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Nate Herk | AI Automation
Every Codex Concept Explained for Non-Coders
A comprehensive tutorial explaining 18 core Codex concepts for non-technical users, organized into four parts covering basics (projects, agents, agent loops, goals), environments (on-premises vs. cloud, working trees), customization (AI models, effort levels, permissions, skills), and scaling tools (browser, sites, subagents, scheduled tasks, voice mode).
Claude Code Mods Are Game Changers. This One Saves Me Money.
Claude Code now supports mods that allow customization of the Cloud desktop app. A user demonstrates a cache management mod that tracks usage data, displays costs, monitors session limits, and provides notifications before cache expiration to help optimize spending.
5 Claude Code Mods That Everyone Needs
The video demonstrates five useful Claude Code modifications (mods) that are plugins enabling customization of the Claude desktop application interface. The creator showcases practical mods including Cache Keeper for token management, Recording Mode for privacy, Goal tracking interface, and Collision Guard for file conflict detection, explaining how anyone can create mods using plain language.
How to Actually Build & Sell Software with AI as a Non-Techie
Dave Fabrikant, a 10+ year Python programmer and AI engineer, discusses how AI coding tools have transformed software development from a specialized skill into an accessible field for non-technical founders. He explains the progression from personal tools to scalable products, emphasizing architecture, security best practices, and the shift from specification-driven to intent-based development.
I Tested Codex's NEW $500/mo Ultrafast mode
A reviewer tests Codex's new $500/month Ultrafast mode, which claims to be 8x faster than standard mode. While results vary by task complexity, Ultrafast delivers significant speed improvements but at substantially higher resource consumption, making it suitable primarily for time-sensitive projects.