How much VRAM does the AMD Radeon RX 7800 XT have?

The AMD Radeon RX 7800 XT has 16 GB of GDDR6 with 624 GB/s memory bandwidth.

What is the AMD Radeon RX 7800 XT best for?

With 16 GB of VRAM, the AMD Radeon RX 7800 XT handles smaller models (7B–14B) at Q4–Q5 quantization, ideal for entry-level local LLM experimentation and lightweight inference.

What LLMs can the AMD Radeon RX 7800 XT run locally?

The AMD Radeon RX 7800 XT can run 46 of the 99 open-weight models tracked by CanItRun natively in VRAM at 8k context. Top options include: Qwen 3.8 27B at Q3_K_M, Ornith 1.5 35B-A3B (MoE) at Q2_K, Ornith 1.5 9B at Q8_0.

Can the AMD Radeon RX 7800 XT run Gemma 4 31B?

Yes. The AMD Radeon RX 7800 XT runs Gemma 4 31B natively in VRAM at Q2_K quantization, achieving approximately 36.2 tokens per second.

Can the AMD Radeon RX 7800 XT run Qwen 3.6 27B?

Yes. The AMD Radeon RX 7800 XT runs Qwen 3.6 27B natively in VRAM at Q3_K_M quantization, achieving approximately 35.1 tokens per second.

Can the AMD Radeon RX 7800 XT run Qwen3 8B?

Yes. The AMD Radeon RX 7800 XT runs Qwen3 8B natively in VRAM at Q8_0 quantization, achieving approximately 50.4 tokens per second.

AMD Radeon RX 7800 XT

The AMD Radeon RX 7800 XT has 16 GB VRAM and 624 GB/s memory bandwidth. It can run 46 of our 99 tracked models natively in VRAM at 8k context.

With 16 GB GDDR6, the AMD Radeon RX 7800 XT is a consumer-tier GPU that can run 46 models natively.

The AMD Radeon RX 7800 XT launched September 6, 2023 at a $499 MSRP as the 16GB RDNA 3 card between the RX 7700 XT and the RX 7900 GRE: a Navi 32 chiplet design with 60 compute units and 3,840 stream processors, 64MB of Infinity Cache, and 16GB of GDDR6 at 19.5 Gbps on a 256-bit bus, which is 624 GB/s. Total board power is 263 W. It has the same capacity as the RTX 4060 Ti 16GB that launched at the same price, with more than double that card's memory bandwidth.

AMD Radeon RX 7800 XT: AMD's 16GB RDNA 3 mid-range card, launched at $499, the same price as NVIDIA's RTX 4060 Ti 16GB, which has the same capacity but 288 GB/s against this card's 624 GB/s, 2.17 times as much. That gap carries almost straight into decode speed on this site, a little compressed by a fixed per-layer cost: Qwen3 14B at Q4_K_M runs 47.0 tok/s here against 23.1 tok/s on the 4060 Ti 16GB. Within AMD's 16GB cards it sits 8% above the RX 7900 GRE's 576 GB/s and 2.6% below the newer RX 9070 XT's 640 GB/s.

46 of this site's 99 tracked models fit natively at 8k context, the same set as every other 16GB card. The strongest are 27B-class dense models at Q3_K_M: Qwen 3.8 27B, Qwen 3.6 27B and UI-Mate 27B each total 15.15 GB against this card's 15.2 GB of usable VRAM and decode at 35.1 tok/s, while Q4_K_M needs 19.02 GB and offloads at 9.7 tok/s. 35B-class mixture-of-experts models such as Ornith 1.5 35B-A3B and the two Qwen 35B-A3B releases fit only at Q2_K (15.12 GB) but decode at 104.3 tok/s, because only about 3B parameters are read per token. Qwen3 14B holds Q4_K_M with room to spare (11.6 GB, 47.0 tok/s), GPT-OSS 20B fits at Q4_K_M (14.55 GB, 115.1 tok/s), and Llama 3.1 8B runs at Q8_0 (10.73 GB, 51.6 tok/s). 70B-class models are out of reach.

AMD's ROCm support tables list the RX 7800 XT (gfx1101) on Linux, limited to Ubuntu 24.04.4 and 22.04.5 or RHEL 10.1 and 9.7, and in the HIP SDK 7.2 table for Windows, so the common advice that ROCm is Linux-only does not apply to this card. llama.cpp's Vulkan backend needs no ROCm install. On the project's public Vulkan llama-bench scoreboard (Llama 2 7B Q4_0; driver not stated), this card generates 118.3 tok/s and reads a 512-token prompt at 2,017 tok/s. This site's decode model (88% of rated bandwidth plus a fixed cost per layer) predicts about 113 tok/s for that 3.56 GiB file, within 5% of the measurement.

Vendor	AMD
Architecture	RDNA 3
VRAM	16 GB
Memory type	GDDR6
Memory bandwidth	624 GB/s
Compute backend	ROCM
Tier	Consumer
Released	2023
Models (native)	46 / 99
Models (offload)	12 / 99

Software: Supported by AMD's ROCm on Linux and by the HIP SDK on Windows (gfx1101). llama.cpp's Vulkan backend also works on both operating systems with no ROCm install.

Need more than 16GB? Rent an RTX 3090 from $0.22/hr ↗

Prices as of Oct 13, 2026. Rates change; check RunPod for current pricing. Affiliate link: CanItRun may earn a commission. Doesn't affect the fit calculation above.

Popular models for this GPU

Qwen 3.5 35B-A3B (MoE)Qwen 3.6 35B-A3B (MoE)Ornith 1.5 35B-A3B (MoE)Nemotron 3 Nano 30B Gemma 4 31B

Models this GPU runs natively in VRAM (46)

Show 41 more

Models that fit with CPU offload (12)

These use system RAM for layers that don't fit in VRAM, so expect much slower inference.

Too large for this GPU (41)

Compare AMD Radeon RX 7800 XT with other GPUs

Continue reading

vram-guides13 min

Best LLMs for 16 GB VRAM GPUs (2026)

Frequently asked questions

How much VRAM does the AMD Radeon RX 7800 XT have?: The AMD Radeon RX 7800 XT has 16 GB of GDDR6 with 624 GB/s memory bandwidth.
What is the AMD Radeon RX 7800 XT best for?: With 16 GB of VRAM, the AMD Radeon RX 7800 XT handles smaller models (7B–14B) at Q4–Q5 quantization, ideal for entry-level local LLM experimentation and lightweight inference.
What LLMs can the AMD Radeon RX 7800 XT run locally?: The AMD Radeon RX 7800 XT can run 46 of the 99 open-weight models tracked by CanItRun natively in VRAM at 8k context. Top options include: Qwen 3.8 27B at Q3_K_M, Ornith 1.5 35B-A3B (MoE) at Q2_K, Ornith 1.5 9B at Q8_0.
Can the AMD Radeon RX 7800 XT run Gemma 4 31B?: Yes. The AMD Radeon RX 7800 XT runs Gemma 4 31B natively in VRAM at Q2_K quantization, achieving approximately 36.2 tokens per second.
Can the AMD Radeon RX 7800 XT run Qwen 3.6 27B?: Yes. The AMD Radeon RX 7800 XT runs Qwen 3.6 27B natively in VRAM at Q3_K_M quantization, achieving approximately 35.1 tokens per second.
Can the AMD Radeon RX 7800 XT run Qwen3 8B?: Yes. The AMD Radeon RX 7800 XT runs Qwen3 8B natively in VRAM at Q8_0 quantization, achieving approximately 50.4 tokens per second.