BREAKING
Grok 4.5 Rated 'Opus-Class' by Fable 5
Output Price per 1M Tokens
Fable 5
50
Opus 4.8
25
Grok 4.5
6
SWE-Bench Pro Scores
Fable 5
80.4
Opus 4.8
69.2
Grok 4.5
64.7
0
%
Grok 4.5 lead
0
$
Grok per task
0
$
Fable per task
Diverging Strategies
Anthropic Fable 5
Mythos-class
●
Strong on long, complex tasks
●
Falls back to Opus 4.8 on risk
●
Higher price, safety-first
xAI Grok 4.5
Coding + agents
●
Token-efficient, low cost
●
Strong for agentic workflows
●
Targets high-volume demand
Performance vs Cost Battle Heats Up
AI NEWS BLITZ
Anthropic's Fable 5 has classified xAI's Grok 4.5 as an Opus-class model.