the perfect Sunday does exist 🌅🥭🐶

🚨 JUST IN: THE US GOVERNMENT JUST SHUT DOWN CLAUDE FABLE 5 FOR THE ENTIRE WORLD > a national security export control directive. effective immediately > every foreign national on earth - including Anthropic's own engineers who built the model - is now locked out > Fable 5 launched 4 days ago > it's already being classified like a weapons system > Anthropic says it's a misunderstanding > the order doesn't care > Opus 4.8 still runs. Fable 5 and Mythos 5 are gone until further notice AI just got its first export control. it won't be the last

Distributed AI Inference at Scale on NVIDIA Dynamo With Gcore and Orange Business x.com/i/broadcasts/1vJpPPVLe…

The Anatomy of a new Claude 'Fable 5' Prompt: 1. Task Start with why, NOT what. Claude 5 connects the dots. 'I'm working on [goal] for [who it's for]. They need [what the output enables]. With that in mind: [task].' 2. Context Files Upload your expertise. Stop explaining in prompts. "Read these files completely before responding: [filename .md] - [what it contains]." The file is the brain. This part never changes. 3. Reference Show Claude 5 what good looks like. "Reference for what I want to achieve: [paste]." One example beats ten instructions. 4. Effort The new change, a few people are talking about. "This is a [routine / hard / hardest-unsolved] problem. Scope it like it's at the top of your range." Teams testing Claude 5 on easy tasks undersell it. Give it your hardest problem. 5. Act "AskUserQuestion" is still the king. Add "When you have enough information to act, act. Don't re-litigate my decisions. While weighing a choice, give a recommendation." 6. Scope Claude 5 over-delivers by default. Control it. "Do the simplest thing that works well. No extra features, refactors, or abstractions. If I'm describing a problem, the deliverable is your assessment." The old one did too little. This one does too much. 7. Delegate One Claude is no longer the limit. "Split independent subtasks across subagents & keep working while they run. Verify with a fresh-context subagent." It's not a chatbot anymore. It's a team lead. 8. Evidence The line that removes fake progress reports. "Before reporting progress, audit every claim against a tool result. If it's unverified, say so. Tests failed? Show the output." Anthropic tested this. It nearly eliminated fabricated status updates. 9. Memory Claude 5 gets smarter every run. If you let it. "Record learnings in [notes .md] — one per file. Update, no duplicate. Delete what turns out wrong." Your prompts expire. Your learning file compounds. 10. Checkpoint It can run for hours. Decide when it stops. "Pause only for: destructive actions, scope changes, or input only I can provide. Never end your turn on a promise." The old fear was Claude stopping too late. The new fear is stopping too early. 11. Report The last block. The first thing you read. "Open with the outcome - the TLDR I'd ask for. Complete sentences. Clear beats short." It worked for hours. You read for ten seconds. Copy the full prompt template download my personal md. files for Claude here: Step 1. Go to how-to-ai.guide. Step 2. Subscribe for free. Don't pay anything. Step 3. Open my welcome email. Step 4. Hit the automatic reply button inside. Step 5. Download my .md files. Ready to upload.

Claude is the one who sets the trend. This creates competition, which leads to progress. @GoogleDeepMind Diffusion Gemma is not standing still and is implementing a new model that introduces an fast approach to text generation. it generates entire blocks of text simultaneously, delivering up to 4x faster text generation on GPUs. Find out how it works 👇

DiffusionGemma is our new experimental open model with up to 4x faster output on dedicated GPUs. Instead of predicting word-by-word, it generates entire blocks of text simultaneously. This lets the model self-correct and format complex markdown in real time.

FABLE 5 PRICING ISN’T EXPENSIVE. BAD TOKEN HYGIENE IS. Claude Fable 5 API pricing: - $10 / 1M input tokens - $50 / 1M output tokens - 1M context window - up to 128k output tokens What this actually means: 1) INPUT IS CHEAP. OUTPUT IS NOT. 1k input tokens = $0.01 1k output tokens = $0.05 So the real cost driver usually isn’t “big context” by itself. It’s: - long answers - huge code generation - repeated full-history chats - multi-step agent loops 2) THE 1M CONTEXT WINDOW DOES NOT MEAN EVERY CALL COSTS $10. You only pay for the tokens you actually send. A 1M-token prompt would cost about $10 on input. But most real requests are far below that. 3) ROUGH COST PER TASK - Short Q&A ~2k input 0.5k output = ~$0.045 - Summarizing an article ~8k input 1k output = ~$0.13 - Analyzing a 10–20 page document ~20k input 2k output = ~$0.30 - Code review / debugging a code chunk ~30k input 3k output = ~$0.45 - Large refactor / agentic coding run ~100k input 8k output = ~$1.40 - Heavy long-context call 1M input 20k output = ~$11.00 - Near-max output call 1M input 128k output = ~$16.40 4) HOW FAR A BUDGET GOES With $10 you roughly get: ~220 short Q&A calls ~77 article summaries ~33 document analyses ~22 code review/debug sessions ~7 big agentic runs With $100 you roughly get: ~2,200 short Q&A calls ~770 article summaries ~330 document analyses ~220 code review/debug sessions ~70 big agentic runs 5) THE REAL KILLER IS CONTEXT ACCUMULATION If you run a 10-step session and each step adds ~5k tokens, you are not paying for just the last 5k. You often keep repaying for the growing conversation history. That means a workflow that “feels small” can quietly become a few hundred thousand input tokens over a session. That’s why agent workflows can get expensive even when each individual prompt looks reasonable. 6) HOW TO KEEP COSTS DOWN - cap output length - avoid sending full chat history every turn - compact/summarize old context - split large docs into chunks - use Fable 5 only for the hardest steps - route simple triage/writing/support to cheaper models - avoid unnecessary retries and tool loops Bottom line: Fable 5 is totally reasonable for high-value work: complex coding, research, long-context analysis, autonomous workflows. But if you use it like a casual chat model and let context/output grow unchecked, you’re not paying for intelligence. You’re paying for sloppy token management.

FABLE 5 PRICING ISN’T EXPENSIVE. BAD TOKEN HYGIENE IS. Claude Fable 5 API pricing: - $10 / 1M input tokens - $50 / 1M output tokens - 1M context window - up to 128k output tokens What this actually means: 1) INPUT IS CHEAP. OUTPUT IS NOT. 1k input tokens = $0.01 1k output tokens = $0.05 So the real cost driver usually isn’t “big context” by itself. It’s: - long answers - huge code generation - repeated full-history chats - multi-step agent loops 2) THE 1M CONTEXT WINDOW DOES NOT MEAN EVERY CALL COSTS $10. You only pay for the tokens you actually send. A 1M-token prompt would cost about $10 on input. But most real requests are far below that. 3) ROUGH COST PER TASK - Short Q&A ~2k input 0.5k output = ~$0.045 - Summarizing an article ~8k input 1k output = ~$0.13 - Analyzing a 10–20 page document ~20k input 2k output = ~$0.30 - Code review / debugging a code chunk ~30k input 3k output = ~$0.45 - Large refactor / agentic coding run ~100k input 8k output = ~$1.40 - Heavy long-context call 1M input 20k output = ~$11.00 - Near-max output call 1M input 128k output = ~$16.40 4) HOW FAR A BUDGET GOES With $10 you roughly get: ~220 short Q&A calls ~77 article summaries ~33 document analyses ~22 code review/debug sessions ~7 big agentic runs With $100 you roughly get: ~2,200 short Q&A calls ~770 article summaries ~330 document analyses ~220 code review/debug sessions ~70 big agentic runs 5) THE REAL KILLER IS CONTEXT ACCUMULATION If you run a 10-step session and each step adds ~5k tokens, you are not paying for just the last 5k. You often keep repaying for the growing conversation history. That means a workflow that “feels small” can quietly become a few hundred thousand input tokens over a session. That’s why agent workflows can get expensive even when each individual prompt looks reasonable. 6) HOW TO KEEP COSTS DOWN - cap output length - avoid sending full chat history every turn - compact/summarize old context - split large docs into chunks - use Fable 5 only for the hardest steps - route simple triage/writing/support to cheaper models - avoid unnecessary retries and tool loops Bottom line: Fable 5 is totally reasonable for high-value work: complex coding, research, long-context analysis, autonomous workflows. But if you use it like a casual chat model and let context/output grow unchecked, you’re not paying for intelligence. You’re paying for sloppy token management.

Introducing Claude Fable 5: a Mythos-class model that we’ve made safe for general use. Its capabilities exceed those of any model we’ve ever made generally available.

Just landed nested subagent support in Claude Code Starting to experiment more with agents kicking off agents as a way to better manage context. Capped at depth=5 to start, going out in today’s release. Lmk what you think!