xAI cloud models
20 models tracked via Artificial Analysis. Compare cloud performance, then find local GGUF versions in the GraySoft model catalog.
| Model | Intelligence | Speed (tok/s) |
|---|---|---|
| Grok 4.3 (high) | 53.2 | 154.875 |
| Grok 4.20 0309 v2 (Reasoning) | 49.3 | 182.827 |
| Grok 4.3 (medium) | 48.8 | 141.11 |
| Grok 4.20 0309 (Reasoning) | 48.5 | 173.416 |
| Grok 4.3 (low) | 43.9 | 120.416 |
| Grok 4 | 41.5 | 0 |
| Grok 4.1 Fast (Reasoning) | 38.6 | 0 |
| Grok 4 Fast (Reasoning) | 35.1 | 0 |
| Grok 3 mini Reasoning (high) | 32.1 | 55.774 |
| Grok 4.3 (Non-reasoning) | 31 | 123.57 |
| Grok 4.20 0309 (Non-reasoning) | 29.7 | 162.108 |
| Grok 4.20 0309 v2 (Non-reasoning) | 29 | 167.136 |
| Grok Code Fast 1 | 28.7 | 0 |
| Grok 3 | 25.2 | 0 |
| Grok 4.1 Fast (Non-reasoning) | 23.6 | 0 |
| Grok 4 Fast (Non-reasoning) | 23.1 | 0 |
| Grok 3 Reasoning Beta | 21.6 | 0 |
| Grok 2 (Dec '24) | 13.9 | 0 |
| Grok Beta | 13.3 | 0 |
| Grok-1 | 11.7 | 0 |
Run models locally with guIDE
Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.