What is the best GPU for running Qwen2.5 0.5B locally?

The cheapest hardware in the UK that runs Qwen2.5 0.5B fully in memory at 64k context is the NVIDIA Tesla P40, £261 at eBay, at about 161 tok/s (estimated). The best value is the NVIDIA GeForce RTX 3080 10GB (about 353 tok/s for £466). The fastest we price is the NVIDIA GeForce RTX 5090, at about 831 tok/s.

Can I run Qwen2.5 0.5B on a 16 GB GPU?

Yes: at Q8_0 it fits fully in 16 GB at 64k context. On the NVIDIA GeForce RTX 5080 it runs at about 445 tok/s (estimated).

Can I run Qwen2.5 0.5B on a 24 GB GPU?

Yes: at Q8_0 it fits fully in 24 GB at 64k context. On the NVIDIA GeForce RTX 3090 Ti it runs at about 468 tok/s (estimated).

How do I run Qwen2.5 0.5B locally?

With Ollama: ollama run hf.co/bartowski/Qwen2.5-0.5B-Instruct-GGUF:Q8_0 (the Q8_0 build). LM Studio and llama.cpp commands are on this page.

What licence is Qwen2.5 0.5B released under?

Apache License 2.0, which allows commercial use.

Qwen2.5 0.5B

494M parameters · 32k context · Apache License 2.0 · released 2024-09-16

chat

In short

The cheapest hardware in the UK that runs Qwen2.5 0.5B fully in memory at 64k context is the NVIDIA Tesla P40, £261 at eBay, at about 161 tok/s (estimated).

Updated 15 October 2026. Speeds are estimates; how they are worked out.

What to run it on in the UK

Fully in memory at 64k context, priced at retailers today.

Cheapest
NVIDIA Tesla P40
161tok/sest.Q8_0
£261
Best value
NVIDIA GeForce RTX 3080 10GB
353tok/sest.Q8_0
£466
Fastest
NVIDIA GeForce RTX 5090
831tok/sest.Q8_0
£5,000
Mac
Mac mini M4 16GB
58tok/sest.Q8_0
£765

All hardware that runs it

Everything we price, best value first. Two cards split the model by layers.

Hardware	Quant	Fits	Decode	Price	tok/s per 1,000
NVIDIA Tesla P40	Q8_0	fits	161 tok/s	£261	615.7
AMD Radeon RX 7600 XT	Q8_0	fits	119 tok/s	£348	340.1
Intel Arc B580	Q8_0	fits	171 tok/s	£349	489.5
NVIDIA GeForce RTX 3060 12GB	Q8_0	fits	167 tok/s	£390	428.2
NVIDIA GeForce RTX 3080 10GB	Q8_0	fits	353 tok/s	£466	757.1
AMD Radeon RX 9060 XT 16GB	Q8_0	fits	132 tok/s	£499	264.1
NVIDIA GeForce RTX 5060 Ti 16GB	Q8_0	fits	208 tok/s	£550	377.8
AMD Radeon RX 7800 XT	Q8_0	fits	257 tok/s	£568	452.2
Intel Arc A770 16GB	Q8_0	fits	210 tok/s	£588	356.3
AMD Radeon RX 9070 XT	Q8_0	fits	263 tok/s	£670	393.3
NVIDIA GeForce RTX 5070	Q8_0	fits	312 tok/s	£695	448.6
Intel Arc Pro B60	Q8_0	fits	171 tok/s	£742	229.9
Mac mini M4 16GB	Q8_0	fits	58 tok/s	£765	76.3
AMD Radeon RX 7900 XT	Q8_0	fits	329 tok/s	£789	417.2
NVIDIA GeForce RTX 4070 Super	Q8_0	fits	234 tok/s	£874	267.6
NVIDIA GeForce RTX 4070 Ti Super	Q8_0	fits	312 tok/s	£950	328.2
NVIDIA GeForce RTX 5070 Ti	Q8_0	fits	416 tok/s	£975	426.4
NVIDIA GeForce RTX 4060 Ti 16GB	Q8_0	fits	134 tok/s	£977	136.7
AMD Radeon RX 7900 XTX	Q8_0	fits	395 tok/s	£999	395.5
NVIDIA GeForce RTX 5080	Q8_0	fits	445 tok/s	£1,199	371.5
Mac mini M4 32GB	Q8_0	fits	58 tok/s	£1,200	48.7
NVIDIA GeForce RTX 4080 Super	Q8_0	fits	342 tok/s	£1,429	238.9
AMD Radeon AI PRO R9700	Q8_0	fits	263 tok/s	£1,674	157.3
NVIDIA GeForce RTX 3090	Q8_0	fits	434 tok/s	£1,707	254.4
NVIDIA GeForce RTX 3090 Ti	Q8_0	fits	468 tok/s	£2,199	212.7
Mac mini M4 Pro 48GB	Q8_0	fits	133 tok/s	£3,000	44.3
GMKtec EVO-X2 Ryzen AI Max+ 395 128GB	Q8_0	fits	105 tok/s	£3,320	31.7
Mac Studio M4 Max 64GB	Q8_0	fits	266 tok/s	£3,399	78.1
NVIDIA GeForce RTX 5090	Q8_0	fits	831 tok/s	£5,000	166.3
Mac Studio M3 Ultra 96GB	Q8_0	fits	398 tok/s	£5,499	72.4
NVIDIA GeForce RTX 4090	Q8_0	fits	468 tok/s	£6,015	77.8
Mac Studio M3 Ultra 256GB	Q8_0	fits	398 tok/s	£8,499	46.9
Mac Studio M3 Ultra 512GB	Q8_0	fits	398 tok/s	£17,495	22.8
NVIDIA RTX PRO 6000 Blackwell	Q8_0	fits	831 tok/s	£29,504	28.2
Mac Studio M4 Max 128GB	Q8_0	fits	266 tok/s	–	–
Mac mini M4 Pro 64GB	Q8_0	fits	133 tok/s	–	–
Framework Desktop Ryzen AI Max+ 395 128GB	Q8_0	fits	105 tok/s	–	–
NVIDIA DGX Spark 128GB	Q8_0	fits	127 tok/s	–	–
ASUS Ascent GX10 128GB	Q8_0	fits	127 tok/s	–	–
Framework Desktop Ryzen AI Max+ 395 64GB	Q8_0	fits	105 tok/s	–	–

Memory needed

Weights, KV cache and runtime buffers at each context length (KV cache at fp16).

Quant	File	@8k	@32k
Q8_0gguf	0.5 GB	1.3 GB	1.8 GB

Run it

Copy a command for the runtime you use.

Q8_0
- Ollamaollama run hf.co/bartowski/Qwen2.5-0.5B-Instruct-GGUF:Q8_0
- LM Studiolms get bartowski/Qwen2.5-0.5B-Instruct-GGUF@q8_0
- llama.cppllama-server -hf bartowski/Qwen2.5-0.5B-Instruct-GGUF:Q8_0

Notes

The card says the full context is 32,768 tokens and generation is up to 8192 tokens. Needs transformers 4.37.0 or later, because older versions raise KeyError: 'qwen2'.

Questions

What is the best GPU for running Qwen2.5 0.5B locally?: The cheapest hardware in the UK that runs Qwen2.5 0.5B fully in memory at 64k context is the NVIDIA Tesla P40, £261 at eBay, at about 161 tok/s (estimated). The best value is the NVIDIA GeForce RTX 3080 10GB (about 353 tok/s for £466). The fastest we price is the NVIDIA GeForce RTX 5090, at about 831 tok/s.
Can I run Qwen2.5 0.5B on a 16 GB GPU?: Yes: at Q8_0 it fits fully in 16 GB at 64k context. On the NVIDIA GeForce RTX 5080 it runs at about 445 tok/s (estimated).
Can I run Qwen2.5 0.5B on a 24 GB GPU?: Yes: at Q8_0 it fits fully in 24 GB at 64k context. On the NVIDIA GeForce RTX 3090 Ti it runs at about 468 tok/s (estimated).
How do I run Qwen2.5 0.5B locally?: With Ollama: ollama run hf.co/bartowski/Qwen2.5-0.5B-Instruct-GGUF:Q8_0 (the Q8_0 build). LM Studio and llama.cpp commands are on this page.
What licence is Qwen2.5 0.5B released under?: Apache License 2.0, which allows commercial use.