The US Department of Commerce lifted export controls on Anthropic's AI models "Claude Fable 5" and "Claude Mythos 5" on June 30, 2026. Access to both models is expected to be restored gradually starting the following day, July 1.
June 30–July 1, 2026 · Anthropic
White House Lifts Export Limits on Anthropic's Frontier Models
Less than three weeks after Commerce imposed national-security export controls, "Claude Fable 5" and "Mythos 5" are cleared again — a compromise between spurring AI innovation and managing risk.
June 12
Export controls imposed on national-security grounds
<3 wks
From controls to full reversal
1M
Token context window supported
The regulatory whiplash — a timeline
Announced → banned → partially released → fully cleared
JUN 9
Fable 5 & Mythos 5 announced
→
JUN 12
Export controls; models paused worldwide
→
JUN 26
Mythos 5 partly released to select orgs
→
JUN 30–JUL 1
Full lifting — Fable 5 included
Mythos-class vs. Opus 4.8 — benchmark leap
Higher is better · new models built from taller stacks
$10 / $50
per 1M input / output tokens — under half of Mythos Preview
up to 90%
input cost cut via prompt caching
Real-world signal
Stripe migrated a 50-million-line Ruby codebase in a single day.
Work estimated at ~2 months for a team — done in one day. GitHub and Cursor rate it state-of-the-art for long-horizon, autonomous coding.
Strengths
Best-in-class agentic, long-horizon coding
"Senior research scientist"–level analysis
Sharp price cut + caching efficiency
Open concerns
Aggressive filters reject harmless queries / downgrade to Opus 4.8
Heavy token consumption and cost
Privacy — 30-day prompt retention
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article…