16GB VRAM

Radeon RX 7800 XT — what LLMs can it run?

63 of 110 indexed models fit comfortably in 16GB at 4K context, each at the highest-quality quant that still leaves headroom.

The short answer for Radeon RX 7800 XT

16 GB, 624 GB/s GDDR6. 63 of 110 models in this index fit comfortably at 4K context; 70 load at all.

Biggest that fits: Mistral Small 24B Instruct — 13.2 GB at AWQ INT4 — 2.8 GB spare at 4K, so longer context comes out of a thin margin.
Room to grow: Stable LM 2 12B Chat — 8.3 GB at Q4_K_M — under 60% of the card, which leaves 7.7 GB for a long context window or a second process.
Speed ceiling: Llama 3.1 8B Instruct at Q4_K_M reads 4.6 GB of weights per token, so 624 GB/s puts a hard ceiling near 135 tok/s. That is arithmetic on two published numbers, not a benchmark — real throughput lands below it. No run on this card has been measured here, so there is nothing to compare the ceiling against.
The wall you will hit: Backend coverage, not capacity. At 624 GB/s the memory system is competitive with NVIDIA cards of the same size; what varies is whether your runtime has a working ROCm build for your kernel and card. llama.cpp and vLLM both ship official ROCm support — this is a setup question, not a hardware ceiling.

Best local LLM for 16GB

32B · 6

Radeon RX 7800 XT — what LLMs can it run? — 32B
Model	Quant	Est. VRAM	Headroom	tok/s
Mistral Small 24B Instruct24B	AWQ INT4	13.23 GB	+2.8 GB	≤55Ceiling
Devstral Small 1.1 24B24B	AWQ INT4	13.02 GB	+3 GB	≤56Ceiling
Magistral Small 1.2 24B24B	AWQ INT4	13.02 GB	+3 GB	≤56Ceiling
Codestral 22B22B	AWQ INT4	12.56 GB	+3.4 GB	≤59Ceiling
ERNIE 4.5 21B-A3B21B-A3B	Q4_K_M	14.05 GB	+1.9 GB	≤361Ceiling
GPT-OSS 20B21B MoE	MXFP4	12.76 GB	+3.2 GB	≤315Ceiling

14B · 18

Radeon RX 7800 XT — what LLMs can it run? — 14B
Model	Quant	Est. VRAM	Headroom	tok/s
InternLM2 20B Chat20B	Q4_K_M	13.43 GB	+2.6 GB	32–42Derived
DeepSeek-Coder-V2-Lite Instruct16B-A2.4B	Q4_K_M	10.08 GB	+5.9 GB	≤452Ceiling
DeepSeek-V2-Lite Chat16B-A2.4B	Q4_K_M	10.08 GB	+5.9 GB	≤452Ceiling
StarCoder2 15B15B	Q4_K_M	10.19 GB	+5.8 GB	42–56Derived
Qwen3 14B Instruct14B	Q5_K_M	11.65 GB	+4.4 GB	36–48Derived
Qwen2.5 14B Instruct14B	Q5_K_M	11.73 GB	+4.3 GB	37–48Derived
DeepSeek-R1-Distill-Qwen-14B14B	Q4_K_M	10.14 GB	+5.9 GB	43–57Derived
Phi-4 14B14B	Q5_K_M	11.77 GB	+4.2 GB	37–48Derived
Phi-3 Medium 14B Instruct14B	Q6_K	12.86 GB	+3.1 GB	33–44Derived
Mistral Nemo 12B Instruct12B	Q6_K	11.14 GB	+4.9 GB	38–51Derived
Gemma 3 12B IT12B	Q5_K_M	9.84 GB	+6.2 GB	44–58Derived
Mellum2.1 12B-A2.5B Thinking12B-A2.5B	Q8_0	13.6 GB	+2.4 GB	≤248Ceiling
Stable LM 2 12B Chat12B	Q4_K_M	8.35 GB	+7.7 GB	52–69Derived
Llama 3.2 11B Vision Instruct11B	Q8_0	12.9 GB	+3.1 GB	33–43Derived
Solar 10.7B Instruct11B	Q4_K_M	7.6 GB	+8.4 GB	59–78Derived
Falcon 3 10B Instruct10B	Q4_K_M	7.28 GB	+8.7 GB	61–80Derived
Gemma 2 9B Instruct9B	Q8_0	11.7 GB	+4.3 GB	39–52Derived
Qwen3.5 9B9B	Q8_0	10.37 GB	+5.6 GB	39–52Derived

7B · 25

Radeon RX 7800 XT — what LLMs can it run? — 7B
Model	Quant	Est. VRAM	Headroom	tok/s
GLM-4-9B-Chat9B	Q8_0	10.16 GB	+5.8 GB	40–53Derived
Qwen3-VL 8B Instruct8B	Q8_0	10.39 GB	+5.6 GB	41–54Derived
Granite 4.2 8B8B	Q8_0	10.45 GB	+5.6 GB	41–54Derived
Ministral 3 8B Instruct8B	Q8_0	10.02 GB	+6 GB	42–56Derived
Qwen2-VL 7B Instruct7B	Q4_K_M	5.49 GB	+10.5 GB	76–101Derived
Granite 3.1 8B Instruct8B	Q4_K_M	5.88 GB	+10.1 GB	77–102Derived
Qwen3 8B Instruct8B	Q6_K	7.64 GB	+8.4 GB	57–75Derived
Llama 3.1 8B Instruct8B	Q8_0	9.47 GB	+6.5 GB	45–59Derived
Nous Hermes 3 Llama 3.1 8B8B	Q4_K_M	5.64 GB	+10.4 GB	79–104Derived
Aya 23 8B8B	Q4_K_M	5.64 GB	+10.4 GB	79–104Derived
OpenChat 3.6 8B8B	Q4_K_M	5.64 GB	+10.4 GB	79–104Derived
DeepSeek-R1-Distill-Llama-8B8B	Q5_K_M	6.51 GB	+9.5 GB	67–89Derived
InternLM2 7B Chat7B	Q4_K_M	5.45 GB	+10.6 GB	81–108Derived
Qwen2.5 7B Instruct7B	Q6_K	6.77 GB	+9.2 GB	61–81Derived
Qwen2.5-Coder 7B Instruct7B	Q4_K_M	5.07 GB	+10.9 GB	83–109Derived
DeepSeek-R1-Distill-Qwen-7B7B	Q4_K_M	5.07 GB	+10.9 GB	83–109Derived
Gemma 4 E4B ITE4B	Q8_0	8.46 GB	+7.5 GB	48–63Derived
OLMo 2 7B Instruct7B	Q8_0	10.3 GB	+5.7 GB	49–65Derived
OLMo 3 7B Instruct7B	Q8_0	10.3 GB	+5.7 GB	49–65Derived
WizardLM-2 7B7B	Q4_K_M	5.14 GB	+10.9 GB	87–115Derived
Mistral 7B Instruct v0.37B	Q6_K	6.75 GB	+9.3 GB	64–85Derived
Zephyr 7B Beta7B	Q6_K	6.75 GB	+9.3 GB	64–85Derived
Gemma 4 E2B ITE2B	Q8_0	5.2 GB	+10.8 GB	≤133Ceiling
Qwen3.5 4B4B	Q8_0	4.97 GB	+11 GB	≤142Ceiling
Gemma 3 4B IT4B	Q8_0	5.05 GB	+11 GB	≤144Ceiling

≤3B · 14

Radeon RX 7800 XT — what LLMs can it run? — ≤3B
Model	Quant	Est. VRAM	Headroom	tok/s
Qwen3 4B Instruct4B	Q6_K	4.05 GB	+12 GB	≤200Ceiling
Phi-4 Mini Instruct3.8B	Q8_0	4.81 GB	+11.2 GB	≤161Ceiling
Phi-3.5 Mini Instruct3.8B	Q8_0	5.89 GB	+10.1 GB	≤162Ceiling
Granite 4.2 3B3B	Q8_0	4.41 GB	+11.6 GB	≤169Ceiling
Llama 3.2 3B Instruct3B	Q8_0	4.05 GB	+12 GB	≤193Ceiling
Qwen2.5 3B Instruct3B	Q8_0	3.59 GB	+12.4 GB	≤200Ceiling
SmolLM3 3B3B	Q8_0	3.73 GB	+12.3 GB	≤201Ceiling
Gemma 2 2B Instruct2B	Q8_0	3.34 GB	+12.7 GB	≤237Ceiling
Qwen3.5 2B2B	Q8_0	2.23 GB	+13.8 GB	≤315Ceiling
Qwen3 1.7B Instruct1.7B	Q8_0	2.39 GB	+13.6 GB	≤359Ceiling
Qwen2.5 1.5B Instruct1.5B	Q8_0	1.83 GB	+14.2 GB	≤403Ceiling
Llama 3.2 1B Instruct1B	Q8_0	1.51 GB	+14.5 GB	≤499Ceiling
Qwen3.5 0.8B0.8B	Q8_0	0.94 GB	+15.1 GB	≤770Ceiling
Qwen2.5 0.5B Instruct0.5B	Q8_0	0.6 GB	+15.4 GB	≤1273Ceiling

How this list is built

Each row is the lowest-perplexity-loss quant of that model whose estimated total — weights plus KV cache at 4K context plus activation buffer — uses at most 88% of the card. That is the calculator's "green" threshold, so every row here has real headroom rather than only just fitting. Raise the context length and the list shortens; the calculator lets you check any combination directly. The speed column is this card: a measured run, a range scaled from this card's own Llama 2 7B benchmark, or a bandwidth ceiling. A figure that only exists for the RTX 4090 is labelled as that card.

A further 7 models load but with no headroom to spare (up to 105% of VRAM) — the Quant Hub’s GPU chips count those too, which is why its number is higher.

Open this card in the VRAM calculator Feel this speed Browse all 110 models

Measured on this card

No benchmark runs in this index were recorded on a Radeon RX 7800 XT. Every figure on this page is calculated from the model architecture and the quant level — treat them as estimates, not measurements.

Cards with the same budget

What fits is decided by memory, so every 16GB card of this type returns the same list. These pages are not different answers — they differ in throughput, which this index does not measure per card.

The same 16GB in system memory is a different machine — everything on this page fits there, at a speed set by your DIMMs rather than a graphics bus: 16 GB RAM (CPU)

Same list, different speed — play one sentence on both cards: Radeon RX 7800 XT vs Radeon RX 9070 XT (Qwen3 14B Instruct) · Radeon RX 7800 XT vs Radeon RX 9070 (Qwen3 14B Instruct) · Radeon RX 7800 XT vs Radeon RX 9060 XT 16G (Qwen3 14B Instruct)

Setup guides for this card

Step-by-step guides written for the Radeon RX 7800 XT's runtime or memory size, with the commands and the checks that prove the model is running on the card.

AMD GPU + llama.cpp via ROCm (Quick Start)

Stepping up

A Radeon RX 7900 XT (20GB) fits 10 more of the indexed models than this card. Radeon RX 7900 XT →

Before buying, hear the speed difference on a model both cards hold: Radeon RX 7800 XT vs Radeon RX 7900 XT (Qwen3 14B Instruct) →

Common questions

What is the best local LLM for a Radeon RX 7800 XT?

For everyday use, Stable LM 2 12B Chat at Q4_K_M — about 8.3 GB of the card's 16 GB at 4K context, leaving 7.7 GB for a longer window. If you want the largest thing that will load, that is Mistral Small 24B Instruct at AWQ INT4 (13.2 GB, 2.8 GB spare). "Best" here means best fit for the memory budget — this index does not run task benchmarks, so it cannot tell you which model is smarter.

Can a Radeon RX 7800 XT run Gemma 4 26B-A4B IT?

Only just. Its smallest build here, Q4_K_M, needs about 16.4 GB against 16.0 GB usable — that loads with nothing else running, with no margin for a longer context window. It is not on the list above, which requires a model to stay inside 88% of usable capacity.

How many tokens per second does a Radeon RX 7800 XT do on an 8B model at Q4?

This index has no measured run on this card, so it does not publish a figure. What can be stated from specifications: generating a token requires reading every weight once, Llama 3.1 8B Instruct at Q4_K_M is 4.6 GB of weights, and this card moves 624 GB/s — a ceiling near 135 tok/s. Batch size, context length, the runtime and how much of the model sits in cache all take you below it.

Radeon RX 7800 XT or RTX 5080 for local LLMs?

They hold the same models: 16 GB against 16 GB fits 63 and 63 of 110 respectively at 4K. The difference is throughput — 624 GB/s against 960 GB/s, a 1.5× gap in how fast the weights can be read, which is what token generation is bound by. The RTX 5080 generates faster on any model both can hold.

This page's figures change when the model or the runtime does.Last updated 2026-10-18 RSS → /feed.xml