BREAKING
xAI launches Grok 4.5
0
$/M
Input tokens
0
$/M
Output tokens
0
K
Context window
Terminal Bench 2.1 scores
Fable 5
84.3
GPT-5.5
83.4
Grok 4.5
83.3
Cost per task in agents
Grok 4.5
2.49
Fable 5
11.8
Planner vs implementer
Fable 5
Orchestrator
●
Plans the work
●
Reviews changes
●
Higher cost
Grok 4.5
Implementer
●
Executes edits
●
High-volume code
●
Low cost
Price-performance shift
AI NEWS BLITZ
xAI has released Grok 4.5, a coding-focused model that matches frontier rivals at a fraction of the cost.