Grok 4
Parameters
-
Context Length
128K
Modality
Text
Architecture
Dense
License
Proprietary
Release Date
1 Jun 2025
Knowledge Cutoff
-
Technical Specifications
Attention Structure
Multi-Head Attention
Hidden Dimension Size
-
Number of Layers
-
Attention Heads
-
Key-Value Heads
-
Activation Function
-
Normalization
-
Position Embedding
Absolute Position Embedding
Grok 4
xAI's powerful conversational model with strong reasoning and coding capabilities. Excellent reasoning (79.13) and language understanding (76.39). Good coding (73.13) and mathematics (83.02) performance. Features real-time knowledge access and direct communication style. Ideal for technical discussions and complex problem-solving.
About Grok
xAI's conversational AI models with real-time knowledge access and strong performance across reasoning, coding, and language tasks. Features extended context windows, fast inference variants, and specialized coding versions. Known for direct communication style and integration with X platform. Includes reasoning variants and optimized versions for different latency requirements.
Other Grok Models
Evaluation Benchmarks
Rank
#19
| Benchmark | Score | Rank |
|---|---|---|
QA Assistant ProLLM QA Assistant | 0.98 | 🥇 1 |
Summarization ProLLM Summarization | 0.98 | 🥉 3 |
StackUnseen ProLLM Stack Unseen | 0.89 | 5 |
Coding Aider Coding | 0.80 | ⭐ 6 |
Reasoning LiveBench Reasoning | 0.79 | ⭐ 8 |
Mathematics LiveBench Mathematics | 0.83 | 12 |
Data Analysis LiveBench Data Analysis | 0.70 | 22 |
Coding LiveBench Coding | 0.73 | 23 |
Graduate-Level QA GPQA | 0.88 | 31 |
Agentic Coding LiveBench Agentic | 0.30 | 32 |
Rankings
Overall Rank
#19
Coding Rank
#26