← CanIRunLocal · All hardware · All models

Can I run BGE-M3 567M locally?

BAAI · 0.567B parameters · embedding.

Ollama weight files: FP16 1.2 GB (bge-m3:567m).

Yes on 56 machines in this list. The smallest is M1 8GB: FP16, about 1.8 GB, via Ollama. ollama pull bge-m3:567m

Setup on DGX Spark 128GB (NVIDIA, ASUS, Dell, Lenovo, HP, MSI, Acer, Gigabyte)

Open BGE-M3 567M in the checker

Hardware that fits

HardwareRuntimeQuantMemoryFitUseCommand
M1 8GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M2 8GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M3 8GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RTX 3060 Ti 8GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RTX 3070 8GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RTX 4060 8GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RX 7600 8GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RTX 3080 10GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M1 16GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M2 16GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M3 16GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M4 16GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RTX 3060 12GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RTX 4070 12GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RTX 5070 12GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RTX 4060 Ti 16GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RTX 4070 Ti Super 16GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RTX 4080 16GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RTX 5080 16GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RX 7800 XT 16GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RX 9070 XT 16GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M2 24GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M3 24GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M4 24GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RTX 3090 24GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RTX 4090 24GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RX 7900 XTX 24GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M1 Pro 32GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M2 Pro 32GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M4 32GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M3 Pro 36GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
Mac Studio M4 Max (32c GPU) 36GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
Mac Studio M5 Max (32c GPU) 36GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
RTX 5090 32GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M4 Pro 48GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
Mac Studio M4 Max (40c GPU) 48GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
Mac Studio M5 Max (40c GPU) 48GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M1 Max 64GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M2 Max 64GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M3 Max 64GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M4 Max 64GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
Mac Studio M4 Max (40c GPU) 64GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
Mac Studio M5 Max (40c GPU) 64GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
DGX Spark 64GB (ASUS, Dell, HP, MSI, Acer, Gigabyte)OllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
Mac Studio M3 Ultra 96GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
Mac Studio M5 Ultra 96GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M3 Max 128GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M4 Max 128GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
Mac Studio M4 Max (40c GPU) 128GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
Mac Studio M5 Max (40c GPU) 128GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
DGX Spark 128GB (NVIDIA, ASUS, Dell, Lenovo, HP, MSI, Acer, Gigabyte)OllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
M2 Ultra 192GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
Mac Studio M3 Ultra 256GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
Mac Studio M5 Ultra (64c GPU) 256GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
Mac Studio M5 Ultra (80c GPU) 256GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m
Mac Studio M3 Ultra (80c GPU) 512GBOllamaFP161.8 GBfits in fast memoryembeddingollama pull bge-m3:567m

Estimates, not benchmarks. Model sizes are ollama.com download sizes; memory needed adds headroom for ~8k tokens of context; speeds are rough figures from memory bandwidth. Data updated 2026-10-18.

A project by OnDevice Labs · llms.txt