RUN THIS LLM
Search local LLM hardware requirements
Command A 111B
Cohere · 111B MoE · General
Latest Cohere MoE model. Strong at multilingual tasks with 23 language support.
VRAM Requirements
| Quantization | VRAM |
|---|---|
| Q4_K_M (smallest) | 66.6 GB |
| Q8_0 (balanced) | 122.1 GB |
| FP16 (full quality) | 222 GB |
Specifications
- Parameters: 111B MoE (27B active per token)
- Category: General
- Max context: 256K tokens
- System RAM: 80 GB minimum
- HuggingFace: CohereForAI/c4ai-command-a-03-2025
Benchmarks
- MMLU: 78 — General knowledge & reasoning (5-shot)
Loading interactive analysis...