command-a-plus-05-2026-w4a4
by CohereLabs
126B params · image-text-to-text · 230 likes · 95.9k downloads
command-a-plus-05-2026-w4a4 is a 126B parameter model. At Q4 quantization it requires 63GB of VRAM. It runs comfortably on H100 SXM5 80 GB (38 tok/s), H200 SXM 141 GB (55 tok/s).
Inference providers
| Provider | $/1M in | $/1M out | Throughput |
|---|---|---|---|
| Cohere | 159 tok/s |
GPU compatibility
| GPU | VRAM | Q4 Decode | Verdict |
|---|---|---|---|
| M4 Max 128GB | 128GB | 3 tok/s | tight |
| A100 PCIe 80 GB | 80GB | 19 tok/s | tight |
| H100 SXM5 80 GB | 80GB | 38 tok/s | comfortable |
| H200 SXM 141 GB | 141GB | 55 tok/s | comfortable |
| M2 Ultra 192GB | 192GB | 5 tok/s | tight |