BREAKING
xAI Ships Grok 4.5 for Coding
Co-Trained With Cursor
1
V9 base model
↓
2
Cursor data
↓
3
Agent behavior
0
$/M
input tokens
0
$/M
output tokens
0
k
context length
0
TPS
generation speed
Snorkel Real-Task Pass Rate
Grok 4.5
29
GPT-5.5
22
Opus 4.8
21
Strengths and Trade-Offs
Grok 4.5
agents
●
High tool-calling reliability
●
Strong token efficiency
●
Long-running tasks
Claude
creative
●
Design and UI
●
Creative work
●
Use-case choice
Cost Efficiency in Focus
AI NEWS BLITZ
xAI has released Grok 4.5, a model built specifically for coding and agentic work.