xAILanguage modelSuperseded
Grok 4
Trained with RL at pre-training scale; set records on HLE and ARC-AGI-2 in mid-2025.
Overall Intelligence
#37
of 56
16.5 pts behind #1
0
races led
200
weeks at #1
72.2
score
- Released
- Jul 9, 2025
- 1.3 yr ago
- Context
- 256K
- 64K max output
- Price
- $3 / $15
- per 1M in/out · $6.00 blended
- Speed
- 70 tok/s
- 9s to first token
- Modalities
- text · image
- outputs text
- Weights
- Closed
- API and apps
The 30-second version
Trained with RL at pre-training scale; set records on HLE and ARC-AGI-2 in mid-2025.
Key takeaways
- 1Trained with RL at pre-training scale; set records on HLE and ARC-AGI-2 in mid-2025.
Where it ranks
Current snapshot · illustrative scores · click any race to watch it
Grok 4Best on each axis
Overall
Coding
40Coding+19.6 pts39Frontend & UI+120 Elo41Backend & APIs+17.6 pts39Distributed Systems+19.5 pts42Full-Stack Apps+20.5 pts38Debugging & Repair+18.7 pts42Agentic Coding+23.3 pts39Code Review & Security+17.6 pts36Data Science & SQL+14.9 pts42Mobile Apps+19.3 pts39Terminal & DevOps+23.9 pts35Competitive Programming+458
Hard Reasoning
Math
Science
Writing & Language
Multimodal
Tool Use & Agents
Speed & Cost
Overall Intelligence over time
#1 frontierGrok 4V4.1 FlashGLM-5.3
Update history
The distilled version first; open any release for the deeper breakdown.
The Grok family