cohere-transcribe-03-2026: hardware requirements
cohere-transcribe-03-2026 needs about 2.3 GB of memory at Q4 quantization — a 1.6 GB download plus room for context and overhead. That fits a machine with 8GB RAM.
Parameters
2.1B
Tier
Runs on a laptop
Licence
See model card
Released
2026-03-27
Will it run on your machine?
8GB RAMRuns well16GB RAMRuns well24GB RAMRuns well32GB RAMRuns well64GB RAMRuns well128GB RAMRuns well8GB VRAMRuns well12GB VRAMRuns well16GB VRAMRuns well24GB VRAMRuns well32GB VRAMRuns well
Download sizes by quantization
Lower quantization means a smaller file and less memory, at some cost to quality. Q4_K_M is the usual starting point. What is this?
| Quant | Download | RAM @ 4K | RAM @ 32K | Source |
|---|---|---|---|---|
| Q4_K_Mrecommended | 1.6 GB | 2.3 GB | 2.3 GB | measured |
| Q5_K_M | 1.8 GB | 2.5 GB | 2.5 GB | measured |
| Q6_K | 2.0 GB | 2.7 GB | 2.7 GB | measured |
| Q8_0 | 2.4 GB | 3.2 GB | 3.2 GB | measured |
| BF16 | 4.1 GB | 5.1 GB | 5.1 GB | measured |
“Measured” sizes are read from published GGUF files (handy-computer/cohere-transcribe-03-2026-gguf). “Estimated” sizes are derived from the parameter count and are typically within a few percent. The RAM columns add the context cache and runtime overhead to the download size.
Model card on HuggingFace →API specs & pricing →Verified 2026-08-09Tracked for 12 months after release