Can BGE M3 run on RTX 5090 32GB?

YES — Runs Great

A75Great

○Estimated from fit model

BGE M3 needs ~6.7 GB VRAM. RTX 5090 32GB has 32.0 GB. With F16 quantization, expect ~8 tok/s.

Runtime: OllamaCapacity: RoomyBandwidth: HighStack: BasicBottleneck: Memory bandwidth

Operating mode

Choose the run profile you care about

Interactive favors responsiveness, while light API and scale-out lean harder on serving readiness. The fit stays the same, but the recommendation lens changes.

Current mode

Balanced

Balanced for general local use. Keeps the ranking neutral across personal and serving workflows.

Capabilities:

Select quantization to explore

F16 (Maximum quality) — 6.7 GB, 8.0 tok/s, Runs well

6.7 GB required32.0 GB available

21% VRAM used

Fit status

Runs well

Decode

8.0 tok/s

TTFT

24346 ms

Safe context

Memory

6.7 GB / 32.0 GB

Memory breakdown

Weights1.2 GB

KV Cache1.1 GB

Runtime1.2 GB

Headroom3.2 GB

See how fast it feels

See how fast it feelsBGE M3 on RTX 5090 32GB
1st promptCold start — includes initialization
>What is local AI inference?
Local AI inference means running an AI language model directly on your own hardware — your laptop, desktop, or server — instead of sending requests to a remote cloud API.

When you run inference locally the model weights are loaded into your GPU or unified memory. Each token you generate requires reading those weights from memory, so memory bandwidth is the main bottleneck for decode speed.

Key benefits of running locally:
 - Full privacy: your prompts never leave your machine
 - No per-token cost or rate limits
 - Works offline once the model is downloaded
 - Latency depends only on your hardware
2nd promptWarm — model ready, no init overhead
>How much VRAM do I need?
It depends on the model size and quantization level. A rough rule of thumb:

 Model size Q4 (4-bit) Q8 (8-bit) FP16
 7B params ~4.3 GB ~7.5 GB ~14 GB
 13B params ~7.9 GB ~13.9 GB ~26 GB
 70B params ~42.7 GB ~74.9 GB ~140 GB

Most people use 4-bit quantization (Q4_K_M) which gives 90-95% of full quality at a fraction of the memory. A 24 GB GPU can comfortably run most 7B-13B models.
Estimated: 8.0 tok/s decode · 24.3s TTFT (warm) · 20 tok/s prefill

What limits this setup

This setup is broadly balanced for this model.

No major red flags

This recommendation has enough memory headroom and acceptable estimated speed for the selected workload.

Best improvement path

Performance by workload

Workload	Grade	Fit	Decode	TTFT	Context
Chat	A	Runs well	8.0 tok/s	13280 ms	8K
Coding	A	Runs well	8.0 tok/s	24346 ms	8K
Agentic Coding	A	Runs well	8.0 tok/s	35412 ms	8K
Reasoning	A	Runs well	8.0 tok/s	28773 ms	8K
RAG	A	Runs well	8.0 tok/s	44266 ms	8K

Quantization options

How BGE M3 (0.5680000185966492B params) fits at each quantization level on RTX 5090 32GB (32.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	0.2 GB	Low	A78
Q3_K_S	3	0.3 GB	Low	A78
NVFP4	4	0.3 GB	Medium	A78
Q4_K_M	4	0.3 GB	Medium	A78
Q5_K_M	5	0.4 GB	High	A78
Q6_K	6	0.5 GB	High	A78
Q8_0	8	0.6 GB	Very High	A79
F16Best for your GPU	16	1.2 GB	Maximum	A79

Get started

Copy-paste commands to run BGE M3 on your machine.

Run

docker run --rm -it ghcr.io/ggerganov/llama.cpp:full \
 --hf-repo "BAAI/bge-m3" \
 --hf-file "bge-m3-F16.gguf" \
 -c 4096 -ngl 99

Your hardware

More models your RTX 5090 32GB can run

Model	Params	Grade	Decode
👁 Alibaba Qwen3-Coder 30B A3B Instruct	30.5B	S	181.6 tok/s
👁 Alibaba Qwen 3.5 27B	27B	S	78.7 tok/s
👁 Alibaba Qwen 3.6 27B	27B	S	79 tok/s
👁 Alibaba Qwen 3.6 35B A3B	35B	S	128.2 tok/s
👁 Alibaba Qwen3-VL 30B A3B Instruct	30B	S	187.8 tok/s

Frequently asked questions

See all results for RTX 5090 32GB See all hardware for BGE M3

URL: https://willitrunai.com/can-run/bge-m3-on-rtx-5090-32gb