Meta cloud models
22 models tracked via Artificial Analysis. Compare cloud performance, then find local GGUF versions in the GraySoft model catalog.
| Model | Intelligence | Speed (tok/s) |
|---|---|---|
| Muse Spark 1.3 (max) | 48.2 | 454.777 |
| Muse Spark 1.3 (xhigh) | 45.2 | 322.972 |
| Muse Spark 1.2 (xhigh) | 39.8 | 250.075 |
| Muse Spark 1.1 (xhigh) | 34.3 | 0 |
| Muse Spark | 31.3 | 0 |
| Muse Glimmer (high) | 18.1 | 102.545 |
| Llama 4 Maverick | 9.3 | 94.704 |
| Llama 3.3 Instruct 70B | 7.7 | 87.22 |
| Llama 3.1 Instruct 405B | 7.3 | 0 |
| Llama 3.1 Instruct 8B | 6.9 | 0 |
| Llama 3.1 Instruct 70B | 6.6 | 0 |
| Llama 4 Scout | 6.5 | 91.683 |
| Llama 3.2 Instruct 90B (Vision) | 6.4 | 0 |
| Llama 3.2 Instruct 3B | 5.7 | 0 |
| Llama 2 Chat 7B | 5.7 | 0 |
| Llama 3 Instruct 70B | 5.5 | 0 |
| Llama 3.2 Instruct 11B (Vision) | 5.4 | 21.308 |
| Llama 2 Chat 70B | 5.3 | 0 |
| Llama 2 Chat 13B | 5.3 | 0 |
| Llama 65B | 5 | 0 |
| Llama 3 Instruct 8B | 4.8 | 0 |
| Llama 3.2 Instruct 1B | 4.8 | 0 |
Run models locally with guIDE
Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.