
None
None
Long Story Video Skill
Hand the model a 10-second to 10-minute story and it manages scripts, prompts and footage from start to finish.

Ads Video Skill
In a single pass, the model writes the script and prompts and assembles footage for a product ad.
3D Science Explainer Video Skill
Convert a tough research finding into an animated 3D explainer
Feedback
freeTrialImage.bannerPity
freeTrialImage.upgradeUnlock
- ✓freeTrialImage.benefitHd
- ✓freeTrialImage.benefitWatermark
- ✓freeTrialImage.benefitUnlimited
Claude Fable 5.2
Claude Fable 5.2 handles complex engineering and research briefs with stronger code comprehension, careful analysis and reliable visual understanding.
All Tools
Discover our comprehensive AI-powered animation toolkit
MiniMax H3
MiniMax H3 AI Video Generator
Seedance 2.5
The Future of AI Video Is Here.

Seedance 2.0
The Future of AI Video Is Here.

Veo3.1
Create Stunning Videos with Veo3.1
FLUX 3 Video Generator

Kling 3.0
Next-Gen AI Video Generator
Grok Video Generator
Create Videos from Text or Images with AI
MiniMax H3 video generator
MiniMax H3 AI Video Generator
Claude Fable 5.2 for Work That Resists Shortcuts
Anthropic built Claude Fable 5.2 for open-ended technical problems that resist quick fixes — shipping large code migrations, or cracking questions physics has left unanswered.
- Code decisions that survive productionMore than nine in ten edits landing in Anthropic's own core codebase now come from Claude; Fable 5.2 cleared long-standing bugs in minutes and ported a critical service from Python to Rust, which now boots about 26x faster.
- Research questions that were still unsolvedWorking through holographic duality, it closed two unresolved problems in black hole physics and formalized the proof of Fermat's Last Theorem within 30 hours.
- Guardrails, with verified access via MythosFable 5.2 and Mythos 5.2 share a single underlying model: the Fable release ships with cyber and biology guardrails so it stays broadly available, while security and biomedical teams request Mythos through the Life Sciences and Cyber Verification Programs.
Claude Fable 5.2: Choosing Tier, Effort and Mode
Line up the model tier, effort setting and operating mode with how demanding the job ahead actually is.
Claude Fable 5.2: Benchmark Scores and Practical Reach
Scores from FrontierCode, Chartography, WANDR, HLE and BenchCAD, multi-day agent orchestration, vulnerability hunting in Mythos 5.2, and lab hardware control via MHS.
Merge-quality scores on FrontierCode
Running at High effort, it posted 45.3% on FrontierCode 2.0 versus Opus 5.5's peak of 45.5% at roughly half the per-task cost, edging past GPT-6 Astra at 42.8%; Xhigh effort lifted the number to 49.1%.
Multi-day agent orchestration
From an empty folder, Fable 5.2 directed over 1,000 subagents across eight days to ship a bootable OS — drivers, file storage, windows, installer — along with a hypervisor supported by more than 3,500 machine-checked proofs.
Finding vulnerabilities with Mythos 5.2
On the internal CVE Rediscovery benchmark spanning 1,829 CVEs, Mythos 5.2 surfaced 10.2% more vulnerabilities than Opus 5.5 and 14.4% more than Mythos 5.1; a security firm found one in six confirmed bugs it flagged had slipped past every other frontier model in the test.
Reading charts and images
At Medium effort it hit 66.4% on Chartography, ahead of Opus 5.5 at Max effort (64.4%) at under a tenth of the per-task cost, and scored 83.5 on BenchCAD 2.1 where Opus 5.5 managed 65.3.
Cheaper data collection and knowledge reasoning
Fable 5.2 matched Opus 5.5's best WANDR score at Medium effort for roughly 40% of the per-task cost, and equaled the top Humanity's Last Exam result at Low effort for less than a tenth of the spend.
Driving lab and factory hardware through MHS
Inside the Model Hardware Standard research preview, one MHS driver built on simple read/write primitives lets agents operate microscopes, liquid handlers and robotic arms concurrently — shrinking integration from weeks to hours across programmable devices and any agent harness over the Model Context Protocol.
Claude Fable 5.2 Questions, Answered
Straight answers on cost, output speed, when to pick which model, pre-release safety testing and how Mythos access works.
How do Fable 5.2 and Mythos 5.2 differ?
They are the same model underneath. Fable 5.2 ships with cyber and biology safeguards so it can be released openly, while Mythos 5.2 is issued to verified security and biomedical teams.
What does Claude Fable 5.2 cost to run?
$8 per million input tokens and $40 per million output tokens. Fast mode is optional and bills at $16 per million input and $80 per million output tokens.
How does its speed compare with Fable 5.1?
Output arrives up to 30% faster than Fable 5.1, and switching on fast mode adds up to 2.5x token generation on top.
When is Fable 5.2 the better pick over Opus 5.5?
Reach for Fable on wide-open, high-ambition work; most everyday tasks are cheaper and quicker on Opus 5.5, so move up or down according to scope and complexity.
Which safety tests ran before launch?
Pre-release testing probed sandbox escapes and other risky behaviors, confirmed alignment on par with recent models, and checked leaner safeguards plus stronger anti-distillation defenses.
Is Fable 5.2 usable for cybersecurity or biomedical work?
Most cybersecurity requests are blocked on Fable 5.2, so that work runs on Mythos through the Cyber Verification Program; Claude Security now ships with Mythos 5.2 included.
Claude Fable 5.2 Is Ready for Your Hardest Problem
Start at $8 per million input and $40 per million output tokens, switch on fast mode whenever latency matters, and request Mythos access through the Cyber and Life Sciences Verification Programs.
