Elon Musk said on X on July 9, 2026 that Anthropic is "obviously currently the leader in AI," and pledged that even as a competitor he would never "cut them off" in a way that badly hurt them.
Jul 9, 2026 · Anthropic
"Obviously the current leader in AI"
A rival's rare endorsement lands as Anthropic ships its Mythos-class models — Claude Fable 5 and the limited-access Mythos 5 — posting state-of-the-art results across nearly every benchmark tested.
10×
acceleration in protein design (life sciences)
50M
lines of code it can handle in one codebase
~80%
of blind tests preferred its novel hypotheses
<5%
of sessions trigger the safety re-routing
Long-horizon performance with persistent memory
Game performance vs. prior Claude Opus 4.8 — 3× higher with persistent notes
API pricing · per million tokens
>10%
gain over Opus 4.8 on some tests
Less than half the price of Claude Mythos Preview.
Two models, one base
Claude Fable 5
Mythos-class model made safe for general use. High-risk queries routed to Opus 4.8.
→
Claude Mythos 5
Same base, safeguards lifted for vetted users via Project Glasswing (cyber defense).
Praise from developers
State of the art on CursorBench
Reliable long-horizon coding autonomy
Stripe: a migration done in one day vs team-months
Spreadsheet work 25–30% faster
Points of friction
Usage and rate limits
Safeguards block some benign queries
Data retention policy changes
Outage from government-mandated suspension
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…