Generally available
Claude API
Use claude-fable-5 and verify the Models API, account eligibility, rate limits, regional support, pricing, and covered-model data controls.
Anthropic model · checked July 30, 2026 Generally available · safeguarded
Quick answer
Fable 5 brings Mythos-class capability to general access.
It is the same underlying model as Mythos 5 with conservative safeguards and fallback behavior. API access is general; Claude subscription treatment depends on the current plan and seat.
Verified provider record
Provider status
Generally available from Anthropic after access was restored globally on July 1, 2026.
Claude API ID
claude-fable-5. Anthropic documents dateless IDs from this generation as pinned snapshots, not evergreen aliases.
Inputs and output
Text and image input, text output, multilingual capability, and vision.
Context and output
1 million token context window and up to 128,000 output tokens in the synchronous Messages API.
Published API price
$10 per million input tokens and $50 per million output tokens. Check current pricing and caching rules.
Flowith access
Not asserted here. Anthropic availability does not establish a Flowith workspace model mapping.
Access map
General availability does not mean one entitlement
Plan-dependent
Claude apps and Code
After July 20, 2026, Max and premium Team or legacy Enterprise seats include limited weekly Fable use; Pro and standard seats use credits from the start.
Provider-dependent
AWS and Google Cloud
Anthropic documents cloud availability, but model IDs, regions, accounts, quotas, contracts, and data routing belong to each platform.
Verify separately
Flowith
Check the live Flowith model selector and account entitlements. Provider release state alone is not evidence of integration.
Safeguard and data boundaries
- Fable 5 and Mythos 5 share the same underlying model.
- Fable uses conservative classifiers for cyber, biology and chemistry, and distillation-related requests.
- When a classifier triggers, Anthropic says the request can fall back to Claude Opus 4.8 and the user is notified.
- Anthropic reported that more than 95% of early Fable sessions involved no fallback, but that is provider launch evidence rather than a workload guarantee.
- Covered-model retention changes can affect organizations using zero-data-retention arrangements; consumer retention is governed separately.
Official sources
- Anthropic launch — shared model, safeguards, launch pricing, capabilities, and data policy.
- Anthropic redeployment update — suspension, restoration, global Fable access, and Mythos restrictions.
- Claude model overview — current IDs, limits, pricing, modalities, cloud surfaces, and versioning.
- Fable 5 plan guide — current plan inclusion and usage-credit rules from July 20.
Continue by decision intent
Plans and usage credits Apply the current July 20 plan rules instead of an expired launch promotion. Fable vs Mythos Compare safeguards, fallback behavior, trusted access, and data boundaries. Claude Mythos 5 Understand the invitation-only Glasswing and trusted-access model record. Claude Opus 5 Consider the generally available enterprise model when Fable safeguards or cost do not fit.
Claude Fable 5 questions, answered
Claude Fable 5 is Anthropic's most capable widely released model as of July 30, 2026. It shares its underlying model with Claude Mythos 5 but adds conservative safeguards for general access.
Anthropic documents claude-fable-5. In this model generation, a dateless Claude model ID is a pinned snapshot rather than an evergreen pointer.
Anthropic documents a 1 million token context window, up to 128,000 output tokens in the synchronous Messages API, and standard API pricing of $10 per million input tokens and $50 per million output tokens.
From July 20, 2026, Anthropic says Pro uses Fable 5 through pay-as-you-go usage credits from the start. Max and premium Team or legacy Enterprise seats can use up to 50% of their shared weekly limit on Fable before credits.
They share the same underlying model and published specifications, but Fable adds conservative safeguards and fallback behavior for broad access. Mythos has safeguards lifted in approved domains and remains restricted to trusted organizations.
This page does not claim a Flowith model mapping. Verify the current Flowith workspace separately from Anthropic's API, apps, plans, and cloud platforms.