Skip to content
RigCheckWhat can I run?

What can a 16 GB graphics card run locally?

What can a 16 GB graphics card run locally? Every answer below is the one RigCheck's own checker for that model gives for each 16 GB card in the finder, with 64 GB of system RAM. Nothing on this page is judged separately, and no answer is a formula from the memory size alone.

With 64 GB of system RAM, across the 21 graphics cards with 16 GB of VRAM in RigCheck's finder, the checkers have an answer for 13 of the 16 models, sizes and routes they cover on at least one of them. Placed wholly or partly on the card, by RigCheck's llama.cpp capacity estimates: Mistral Medium 3.5 in llama.cpp (4 of 21), Qwen3.8-27B in llama.cpp (17 of 21, without caveats or conditions on 8), Gemma 4 31B (17 of 21, without caveats or conditions on 8), Gemma 4 26B-A4B (17 of 21, without caveats or conditions on 8), Gemma 4 12B (17 of 21, without caveats or conditions on 8), Gemma 4 E4B (17 of 21, without caveats or conditions on 8), Gemma 4 E2B (17 of 21, without caveats or conditions on 8). On the processor in system RAM alone, the card unused: Qwen3.8-27B in llama.cpp (4 of 21), Gemma 4 31B (4 of 21), Gemma 4 26B-A4B (4 of 21), Gemma 4 12B (4 of 21), Gemma 4 E4B (4 of 21), Gemma 4 E2B (4 of 21). Answers that name no placement (Strata's published requirements, published route and recipe matches, planning budgets; some rest on system RAM alone, as the table says): Qwen3.8-Flash-Next in the Strata engine (all 21), Qwen-Image-2.1 (image generation) (3 of 21), MiniMax H3 (video + audio) (7 of 21), Nemotron 3.5 Bolt 74B-A9B (17 of 21), Nemotron 3.5 Super VL 120B-A12B (17 of 21), FLUX 3 Image Klein (int8, ComfyUI) (8 of 21). Which cards get an answer, and which answer, depends on the card and not only on its memory: the table lists them card by card. The llama.cpp language-model answers are capacity estimates and the Nemotron 3.5, GLM-5.3-Flash and DeepSeek V4.1 Flash answers are published route matches, not measured setups.

16 GB of VRAM: model by model

Language models

Model (checker)With 64 GB RAM, card by card
Qwen3.8-Flash-Next in the Strata engine
✓ Runs
IQ3_S
Fits by the docs' rule (experts + ~10 GB), with little room to spare: close the browser while it runs.
On 11 of 21 cards: GeForce RTX 5080, GeForce RTX 5070 Ti, GeForce RTX 5060 Ti, GeForce RTX 4080 / 4080 Super, GeForce RTX 4070 Ti Super, GeForce RTX 4060 Ti, Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT
✓ Runs
IQ3_S
Fits by the docs' rule (experts + ~10 GB), with little room to spare: close the browser while it runs. RX 6800 / 6900 series support is community-reported; Strata 0.2.1 has no ready-made engine for it, so you build it yourself (docs/AMD_HIP.md).
On 4 of 21 cards: Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT
✓ Runs
IQ3_S
Fits by the docs' rule (experts + ~10 GB), with little room to spare: close the browser while it runs. Laptop GPUs run at lower power than desktop cards; expect lower speeds.
On 1 of 21 cards: GeForce RTX 4090 Laptop
! Experimental path
IQ3_S
Linux only
Fits by the docs' rule (experts + ~10 GB), with little room to spare: close the browser while it runs. Covered by the community "Older GPUs" guide (Linux). The published MI50 benchmark used two cards. Its gfx906 build failed from engine 0.1.44 to 0.2.0; 0.2.1 builds again but has not been run on a card (AMD_HIP.md).
On 2 of 21 cards: Radeon VII / Instinct MI50
! Experimental path
IQ3_S
Fits by the docs' rule (experts + ~10 GB), with little room to spare: close the browser while it runs. Pascal / Volta cards use Strata's separate CUDA 12 engine (the cuda12 app on Windows or Linux, or a Linux build), covered by the community "Older GPUs" guide.
On 1 of 21 cards: Tesla V100
! Experimental path
IQ3_S
Fits by the docs' rule (experts + ~10 GB), with little room to spare: close the browser while it runs. Not on Strata's supported list, though setup accepts it (unvalidated). Its chip (gfx1102) is in the ready-made RX 7000 engines (Windows rdna3, Linux rx7900), which the maintainers have not run on an RX 7000 card. A community Linux Swift IQ3_XXS run is documented in AMD_HIP.md; that report is not an IQ2_XS benchmark.
On 1 of 21 cards: Radeon RX 7600 XT
! Experimental path
IQ3_S
Linux only
Fits by the docs' rule (experts + ~10 GB), with little room to spare: close the browser while it runs. Intel Arc runs through Strata's experimental Intel (SYCL) engine, which you build from source on Linux (the app and the release zips have no Intel engine).
On 1 of 21 cards: Intel Arc A770
Mistral Medium 3.5 in llama.cpp
! Fits with caveats (capacity estimate)
UD-IQ3_XXS · GPU and system RAM (16 of 88 layers on the GPU): a dense model runs slowly from RAM
RigCheck capacity estimate, not a measured setup.
On 4 of 21 cards: GeForce RTX 5080, GeForce RTX 4080 / 4080 Super, GeForce RTX 4060 Ti
– Not covered
On 17 of 21 cards: GeForce RTX 5070 Ti, GeForce RTX 5060 Ti, GeForce RTX 4070 Ti Super, GeForce RTX 4090 Laptop, Tesla V100, Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT, Radeon RX 7600 XT, Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT, Radeon VII / Instinct MI50, Intel Arc A770
Qwen3.8-27B in llama.cpp
✓ Fits (capacity estimate)
UD-IQ3_XXS · GPU memory (input embeddings in system RAM)
On this Radeon card this is a capacity estimate only: we found no source for this model with llama.cpp's own AMD builds. Ollama lists this card for its ROCm support, but its default package (ollama run qwen3.8:27b, 18 GB) is a different file from the one checked here, so compare its size with your memory first.
RigCheck capacity estimate, not a measured setup.
On 9 of 21 cards: Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT, Radeon RX 7600 XT, Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT
✓ Fits (capacity estimate)
UD-IQ3_XXS · GPU memory (input embeddings in system RAM)
RigCheck capacity estimate, not a measured setup.
On 8 of 21 cards: GeForce RTX 5080, GeForce RTX 5070 Ti, GeForce RTX 5060 Ti, GeForce RTX 4080 / 4080 Super, GeForce RTX 4070 Ti Super, GeForce RTX 4060 Ti, GeForce RTX 4090 Laptop
! Fits with caveats (capacity estimate)
Q8_0 · system RAM, CPU only
RigCheck capacity estimate, not a measured setup.
On 4 of 21 cards: Tesla V100, Radeon VII / Instinct MI50, Intel Arc A770
Gemma 4 (E2B, E4B, 12B, 26B-A4B, 31B) in llama.cpp
✓ Best size: fits (capacity estimate)
Gemma 4 31B: fits (capacity estimate) (UD-IQ2_XXS, GPU memory (input embeddings in system RAM)); Gemma 4 26B-A4B: fits (capacity estimate) (UD-IQ3_XXS, GPU memory (input embeddings in system RAM)); Gemma 4 12B: fits (capacity estimate) (Q6_K, GPU memory (input embeddings in system RAM)); Gemma 4 E4B: fits (capacity estimate) (BF16, GPU memory (input embeddings in system RAM)); Gemma 4 E2B: fits (capacity estimate) (BF16, GPU memory (input embeddings in system RAM))
On this Radeon card these are capacity estimates only: we found no source for Gemma 4 with llama.cpp's own AMD builds. Ollama lists this card for its ROCm support, but its default packages (gemma4:31b 19–20 GB, gemma4:26b 16–19 GB, gemma4:12b 7.7–8.0 GB, gemma4:e4b 6.6–9.5 GB, gemma4:e2b 4.6–7.5 GB) are different files from the ones checked here.
RigCheck capacity estimate, not a measured setup.
On 9 of 21 cards: Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT, Radeon RX 7600 XT, Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT
✓ Best size: fits (capacity estimate)
Gemma 4 31B: fits (capacity estimate) (UD-IQ2_XXS, GPU memory (input embeddings in system RAM)); Gemma 4 26B-A4B: fits (capacity estimate) (UD-IQ3_XXS, GPU memory (input embeddings in system RAM)); Gemma 4 12B: fits (capacity estimate) (Q6_K, GPU memory (input embeddings in system RAM)); Gemma 4 E4B: fits (capacity estimate) (BF16, GPU memory (input embeddings in system RAM)); Gemma 4 E2B: fits (capacity estimate) (BF16, GPU memory (input embeddings in system RAM))
RigCheck capacity estimate, not a measured setup.
On 8 of 21 cards: GeForce RTX 5080, GeForce RTX 5070 Ti, GeForce RTX 5060 Ti, GeForce RTX 4080 / 4080 Super, GeForce RTX 4070 Ti Super, GeForce RTX 4060 Ti, GeForce RTX 4090 Laptop
! Best size: fits with caveats (capacity estimate)
Gemma 4 31B: fits with caveats (capacity estimate) (Q8_0, system RAM, CPU only); Gemma 4 26B-A4B: fits with caveats (capacity estimate) (BF16, system RAM, CPU only); Gemma 4 12B: fits with caveats (capacity estimate) (BF16, system RAM, CPU only); Gemma 4 E4B: fits with caveats (capacity estimate) (BF16, system RAM, CPU only); Gemma 4 E2B: fits with caveats (capacity estimate) (BF16, system RAM, CPU only)
RigCheck capacity estimate, not a measured setup.
On 4 of 21 cards: Tesla V100, Radeon VII / Instinct MI50, Intel Arc A770
Nemotron 3.5 (Bolt 74B-A9B, Super VL 120B-A12B)
! Best size: a published route or recipe matches, with caveats
Nemotron 3.5 Bolt 74B-A9B: a published route or recipe matches, with caveats (Matches Unsloth's short-context budget: about 32 GB of RAM or unified memory); Nemotron 3.5 Super VL 120B-A12B: a published route or recipe matches, with caveats (Matches Unsloth's 64 GB budget: UD-IQ1_M (46.1 GB) or UD-IQ2_XXS (51.1 GB))
Nemotron 3.5 Bolt 74B-A9B: We found no source covering AMD GPUs for this model, so only system RAM is counted here. Unsloth budgets around 32 GB of RAM or unified memory for short-context use (48–64 GB for more headroom) with its 22 GB IQ2_XXS GGUF in llama.cpp; longer contexts, images and video need more. This system RAM is 64 GB. That is Unsloth's planning estimate, not a measured run on this hardware. Nemotron 3.5 Super VL 120B-A12B: We found no source covering AMD or Intel GPUs for this model, so only system RAM is counted. Unsloth's planning estimate for a modest context, including the OS, runtime, KV cache and vision file; not a measured run. This setup counts 64 GB of RAM = 64 GB. Longer contexts and video need more.
On 9 of 21 cards: Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT, Radeon RX 7600 XT, Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT
! Best size: a published route or recipe matches, with caveats
Nemotron 3.5 Bolt 74B-A9B: a published route or recipe matches, with caveats (Matches Unsloth's short-context budget: about 32 GB of RAM or unified memory); Nemotron 3.5 Super VL 120B-A12B: a published route or recipe matches, with caveats (Matches Unsloth's 80 GB budget: UD-IQ2_M, UD-Q2_K_XL or UD-IQ3_XXS (55.9–59.8 GB))
Nemotron 3.5 Bolt 74B-A9B: NVIDIA publishes no recipe for a 16 GB card; its discrete-GPU recipes start at 32 GB. Unsloth budgets around 32 GB of RAM or unified memory for short-context use (48–64 GB for more headroom) with its 22 GB IQ2_XXS GGUF in llama.cpp; longer contexts, images and video need more. This system RAM is 64 GB. That is Unsloth's planning estimate, not a measured run on this hardware. Nemotron 3.5 Super VL 120B-A12B: Unsloth's planning estimate for a modest context, including the OS, runtime, KV cache and vision file; not a measured run. This setup counts 16 GB of VRAM + 64 GB of RAM = 80 GB. Longer contexts and video need more.
On 8 of 21 cards: GeForce RTX 5080, GeForce RTX 5070 Ti, GeForce RTX 5060 Ti, GeForce RTX 4080 / 4080 Super, GeForce RTX 4070 Ti Super, GeForce RTX 4060 Ti, GeForce RTX 4090 Laptop
– Not covered
On 4 of 21 cards: Tesla V100, Radeon VII / Instinct MI50, Intel Arc A770
No answer on any of these cards: Qwen3.8-Flash-Next in llama.cpp, Mistral Large 4, Kolibri-1 (official FP8 route), GLM-5.3-Flash (Unsloth GGUF), DeepSeek V4.1 Flash (September release, 1017 update) in DwarfStar
Model (checker)With 64 GB RAM, card by card
Qwen3.8-Flash-Next in llama.cpp
✕ Does not fit
Even the smallest file (UD-IQ1_S, 72.5 GB) needs about 76.1 GB in system RAM. This checker does not cover your graphics card, so it is not counted; its memory could change the answer.
On 15 of 21 cards: GeForce RTX 4070 Ti Super, GeForce RTX 4090 Laptop, Tesla V100, Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT, Radeon RX 7600 XT, Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT, Radeon VII / Instinct MI50, Intel Arc A770
✕ Does not fit
Even the smallest file (UD-IQ1_S, 72.5 GB) needs about 76.1 GB.
On 6 of 21 cards: GeForce RTX 5080, GeForce RTX 5070 Ti, GeForce RTX 5060 Ti, GeForce RTX 4080 / 4080 Super, GeForce RTX 4060 Ti
Mistral Large 4
… Weights not released
smallest estimate ~210 GB vs 80 GB of memory (VRAM + RAM, or a Mac's unified memory)
On all 21 cards
Kolibri-1 (official FP8 route)
– Not covered
On all 21 cards
GLM-5.3-Flash (Unsloth GGUF)
✕ Does not fit
Unsloth's table puts the 1-bit GGUF (93.1–97.6 GB) at 100 GB of total memory, RAM + VRAM or unified memory. This setup counts 16 GB of VRAM + 64 GB of RAM = 80 GB.
On 7 of 21 cards: GeForce RTX 5080, GeForce RTX 5070 Ti, GeForce RTX 5060 Ti, GeForce RTX 4080 / 4080 Super, GeForce RTX 4070 Ti Super, GeForce RTX 4060 Ti
✕ Does not fit
Unsloth's table puts the 1-bit GGUF (93.1–97.6 GB) at 100 GB of total memory, RAM + VRAM or unified memory. This setup counts 64 GB of RAM (this card is not counted: Unsloth names CUDA, CPU-only and Metal builds, no AMD or Intel GPU build).
On 1 of 21 cards: Radeon RX 9070 XT
– Not covered
On 13 of 21 cards: GeForce RTX 4090 Laptop, Tesla V100, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT, Radeon RX 7600 XT, Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT, Radeon VII / Instinct MI50, Intel Arc A770
DeepSeek V4.1 Flash (September release, 1017 update) in DwarfStar
– Not covered
On all 21 cards

Image models

Model (checker)With 64 GB RAM, card by card
Qwen-Image-2.1 (image generation)
! Planning estimate only
Above the all-components-resident planning budget; offload peak unverified
Uncalibrated allowances, not a minimum or a guarantee. Full checkpoint files are counted in host RAM even when mmap or sequential loading can use less. Actual peak depends on attention, prefix caches, VAE tiling, reference sizes and backend. No speed prediction.
On 3 of 21 cards: GeForce RTX 5080, GeForce RTX 5070 Ti, GeForce RTX 5060 Ti
– Not covered
On 18 of 21 cards: GeForce RTX 4080 / 4080 Super, GeForce RTX 4070 Ti Super, GeForce RTX 4060 Ti, GeForce RTX 4090 Laptop, Tesla V100, Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT, Radeon RX 7600 XT, Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT, Radeon VII / Instinct MI50, Intel Arc A770
FLUX 3 Image Klein (int8, ComfyUI)
! A published route or recipe matches, with caveats
Matches BFL's 8 GB guidance: the int8 files in ComfyUI
BFL: the int8 ComfyUI files run on 8 GB GPUs such as the RTX 3070. Image model, int8 text encoder and VAE together are 11.0 GB; by our reading, on a card smaller than that ComfyUI keeps them on the GPU one at a time and needs system RAM for the rest. BFL gives no RAM figure. The prompt upsampler BFL says not to skip is a 9B model (6.72 GB w4a8, 9.75 GB int8 or 18.8 GB BF16 in ComfyUI); BFL does not say whether its 8 GB figure includes it. When it does not fit with the image model, run it on its own first (BFL's README) or switch it off in the ComfyUI template, with weaker results.
On 8 of 21 cards: GeForce RTX 5080, GeForce RTX 5070 Ti, GeForce RTX 5060 Ti, GeForce RTX 4080 / 4080 Super, GeForce RTX 4070 Ti Super, GeForce RTX 4060 Ti, GeForce RTX 4090 Laptop
– Not covered
On 13 of 21 cards: Tesla V100, Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT, Radeon RX 7600 XT, Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT, Radeon VII / Instinct MI50, Intel Arc A770

Video models

Model (checker)With 64 GB RAM, card by card
MiniMax H3 (video + audio)
! A published route or recipe matches, with caveats
Hardware memory matches a published video recipe; exact clip peak unverified
H3-Base FL2VA, source recipe canvases only. Disk-streaming routes assume an NVMe drive. Host-RAM and lab assumptions remain in the full checker; full 2K needs an API.
Matching recipes: SGLang Recipe A (lab): one card, DiT, text encoder and VAE streamed from the checkpoint file on an NVMe drive (Recipe A), BF16; measured at 864×480 (at the native 768P the cookbook says to use video_vae=24 if the decode runs out of memory, not measured); lab run with VRAM capped at 12 GiB and a simulated 32 GB host: shows the memory fits, not your card's speed; a real 32 GB machine re-reads the weights from disk on every step
On 7 of 21 cards: GeForce RTX 5080, GeForce RTX 5070 Ti, GeForce RTX 5060 Ti, GeForce RTX 4080 / 4080 Super, GeForce RTX 4070 Ti Super, GeForce RTX 4060 Ti
– Not covered
On 14 of 21 cards: GeForce RTX 4090 Laptop, Tesla V100, Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT, Radeon RX 7600 XT, Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT, Radeon VII / Instinct MI50, Intel Arc A770

Quick setup, as in the home page finder: one card, a 32K context and no image input for the llama.cpp capacity estimates (Nemotron 3.5, GLM-5.3-Flash and DeepSeek V4.1 Flash are matched against published recipes, memory tables and routes at their own context instead (NVIDIA's 64K / 256K recipes; Unsloth's GLM table states no context length; DwarfStar's routes per DeepSeek checkpoint (documented for the September weights, experimental for 1017), not llama.cpp).), the better of Windows and Linux for Strata, an NVMe SSD, 64 GB of system RAM; for the image and video rows, Qwen-Image (a planning budget): base model, INT8, 512×512, batch one, no reference images; Kolibri-1: the official FP8 route at a 32K context; MiniMax H3: local H3-Base FL2VA recipes, assuming an NVMe drive; Nemotron 3.5: Bolt and Super VL on one card with the RAM given; FLUX 3 Image Klein: Black Forest Labs' int8 ComfyUI route on the card. "With caveats" means part of the model sits in system RAM (slower), an unmeasured peak, or a route the checker marks conditional. Cards with the same answer, word for word, share a line; "RTX 3070 / 3070 Ti" is two cards the checkers treat alike. There is no speed line, because speeds differ between cards with the same memory; where a checker's own conditions mention a speed, they name the hardware it was measured on. Each card's page and the full checkers show measured speeds, labeled. Pick your card in a checker for your exact setup.

What system RAM changes

The checkers' own answers for the same cards with 32 GB instead of 64 GB of system RAM, and, for 64 GB, the smallest larger RAM size for which a checker gives a better answer. Parts of a large model that do not fit on the card are kept in system RAM, so RAM matters as much as VRAM for the big mixture-of-experts models.

With 32 GB of RAM

With more than 64 GB of RAM

The 16 GB cards on this page

21 cards in 17 finder entries; an entry such as "RTX 3070 / 3070 Ti" is two cards the checkers treat alike. Each card's memory as its vendor states it (read 2026-10-18):

A Mac's unified memory is shared by the graphics and the processor, so Macs have their own pages.

FAQ

Can a 16 GB graphics card run Qwen3.8-Flash-Next in the Strata engine?

In RigCheck's quick setup with 64 GB of system RAM, the checker has an answer on all 21 16 GB cards in the finder. On 11 of 21 cards: GeForce RTX 5080, GeForce RTX 5070 Ti, GeForce RTX 5060 Ti, GeForce RTX 4080 / 4080 Super, GeForce RTX 4070 Ti Super, GeForce RTX 4060 Ti, Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT: Yes: IQ3_S; Fits by the docs' rule (experts + ~10 GB), with little room to spare: close the browser while it runs. On 4 of 21 cards: Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT: Yes: IQ3_S; Fits by the docs' rule (experts + ~10 GB), with little room to spare: close the browser while it runs. RX 6800 / 6900 series support is community-reported; Strata 0.2.1 has no ready-made engine for it, so you build it yourself (docs/AMD_HIP.md). On 1 of 21 cards: GeForce RTX 4090 Laptop: Yes: IQ3_S; Fits by the docs' rule (experts + ~10 GB), with little room to spare: close the browser while it runs. Laptop GPUs run at lower power than desktop cards; expect lower speeds. On 2 of 21 cards: Radeon VII / Instinct MI50: Only on an experimental path: IQ3_S; Linux only; Fits by the docs' rule (experts + ~10 GB), with little room to spare: close the browser while it runs. Covered by the community "Older GPUs" guide (Linux). The published MI50 benchmark used two cards. Its gfx906 build failed from engine 0.1.44 to 0.2.0; 0.2.1 builds again but has not been run on a card (AMD_HIP.md). On 1 of 21 cards: Tesla V100: Only on an experimental path: IQ3_S; Fits by the docs' rule (experts + ~10 GB), with little room to spare: close the browser while it runs. Pascal / Volta cards use Strata's separate CUDA 12 engine (the cuda12 app on Windows or Linux, or a Linux build), covered by the community "Older GPUs" guide. On 1 of 21 cards: Radeon RX 7600 XT: Only on an experimental path: IQ3_S; Fits by the docs' rule (experts + ~10 GB), with little room to spare: close the browser while it runs. Not on Strata's supported list, though setup accepts it (unvalidated). Its chip (gfx1102) is in the ready-made RX 7000 engines (Windows rdna3, Linux rx7900), which the maintainers have not run on an RX 7000 card. A community Linux Swift IQ3_XXS run is documented in AMD_HIP.md; that report is not an IQ2_XS benchmark. On 1 of 21 cards: Intel Arc A770: Only on an experimental path: IQ3_S; Linux only; Fits by the docs' rule (experts + ~10 GB), with little room to spare: close the browser while it runs. Intel Arc runs through Strata's experimental Intel (SYCL) engine, which you build from source on Linux (the app and the release zips have no Intel engine). The full checker takes your exact setup (RAM, disk, operating system, processor) and gives the install steps.

Can a 16 GB graphics card run Mistral Medium 3.5 in llama.cpp?

In RigCheck's quick setup with 64 GB of system RAM, the checker has an answer on 4 of the 21 16 GB cards in the finder: on 4 of them the model is placed wholly or partly on the card. On 4 of 21 cards: GeForce RTX 5080, GeForce RTX 4080 / 4080 Super, GeForce RTX 4060 Ti: With caveats, by RigCheck's capacity estimate: UD-IQ3_XXS · GPU and system RAM (16 of 88 layers on the GPU): a dense model runs slowly from RAM; RigCheck capacity estimate, not a measured setup. No answer on: GeForce RTX 5070 Ti, GeForce RTX 5060 Ti, GeForce RTX 4070 Ti Super, GeForce RTX 4090 Laptop, Tesla V100, Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT, Radeon RX 7600 XT, Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT, Radeon VII / Instinct MI50, Intel Arc A770. The full checker takes your exact setup (context length, RAM, image input) and gives the command to start it.

Can a 16 GB graphics card run Qwen3.8-27B in llama.cpp?

In RigCheck's quick setup with 64 GB of system RAM, the checker has an answer on all 21 16 GB cards in the finder: on 17 of them the model is placed wholly or partly on the card, on 4 it runs on the processor in system RAM alone. On 9 of 21 cards: Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT, Radeon RX 7600 XT, Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT: Yes, by RigCheck's capacity estimate: UD-IQ3_XXS · GPU memory (input embeddings in system RAM); On this Radeon card this is a capacity estimate only: we found no source for this model with llama.cpp's own AMD builds. Ollama lists this card for its ROCm support, but its default package (ollama run qwen3.8:27b, 18 GB) is a different file from the one checked here, so compare its size with your memory first; RigCheck capacity estimate, not a measured setup. On 8 of 21 cards: GeForce RTX 5080, GeForce RTX 5070 Ti, GeForce RTX 5060 Ti, GeForce RTX 4080 / 4080 Super, GeForce RTX 4070 Ti Super, GeForce RTX 4060 Ti, GeForce RTX 4090 Laptop: Yes, by RigCheck's capacity estimate: UD-IQ3_XXS · GPU memory (input embeddings in system RAM); RigCheck capacity estimate, not a measured setup. On 4 of 21 cards: Tesla V100, Radeon VII / Instinct MI50, Intel Arc A770: With caveats, by RigCheck's capacity estimate: Q8_0 · system RAM, CPU only; RigCheck capacity estimate, not a measured setup. The full checker takes your exact setup (context length, RAM, image input) and gives the command to start it.

Can a 16 GB graphics card run Gemma 4 (E2B, E4B, 12B, 26B-A4B, 31B) in llama.cpp?

In RigCheck's quick setup with 64 GB of system RAM, the checker has an answer on all 21 16 GB cards in the finder: on 17 of them the model is placed wholly or partly on the card, on 4 it runs on the processor in system RAM alone. On 9 of 21 cards: Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT, Radeon RX 7600 XT, Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT: Yes, by RigCheck's capacity estimate: Gemma 4 31B: fits (capacity estimate) (UD-IQ2_XXS, GPU memory (input embeddings in system RAM)); Gemma 4 26B-A4B: fits (capacity estimate) (UD-IQ3_XXS, GPU memory (input embeddings in system RAM)); Gemma 4 12B: fits (capacity estimate) (Q6_K, GPU memory (input embeddings in system RAM)); Gemma 4 E4B: fits (capacity estimate) (BF16, GPU memory (input embeddings in system RAM)); Gemma 4 E2B: fits (capacity estimate) (BF16, GPU memory (input embeddings in system RAM)); On this Radeon card these are capacity estimates only: we found no source for Gemma 4 with llama.cpp's own AMD builds. Ollama lists this card for its ROCm support, but its default packages (gemma4:31b 19–20 GB, gemma4:26b 16–19 GB, gemma4:12b 7.7–8.0 GB, gemma4:e4b 6.6–9.5 GB, gemma4:e2b 4.6–7.5 GB) are different files from the ones checked here; RigCheck capacity estimate, not a measured setup. On 8 of 21 cards: GeForce RTX 5080, GeForce RTX 5070 Ti, GeForce RTX 5060 Ti, GeForce RTX 4080 / 4080 Super, GeForce RTX 4070 Ti Super, GeForce RTX 4060 Ti, GeForce RTX 4090 Laptop: Yes, by RigCheck's capacity estimate: Gemma 4 31B: fits (capacity estimate) (UD-IQ2_XXS, GPU memory (input embeddings in system RAM)); Gemma 4 26B-A4B: fits (capacity estimate) (UD-IQ3_XXS, GPU memory (input embeddings in system RAM)); Gemma 4 12B: fits (capacity estimate) (Q6_K, GPU memory (input embeddings in system RAM)); Gemma 4 E4B: fits (capacity estimate) (BF16, GPU memory (input embeddings in system RAM)); Gemma 4 E2B: fits (capacity estimate) (BF16, GPU memory (input embeddings in system RAM)); RigCheck capacity estimate, not a measured setup. On 4 of 21 cards: Tesla V100, Radeon VII / Instinct MI50, Intel Arc A770: With caveats, by RigCheck's capacity estimate: Gemma 4 31B: fits with caveats (capacity estimate) (Q8_0, system RAM, CPU only); Gemma 4 26B-A4B: fits with caveats (capacity estimate) (BF16, system RAM, CPU only); Gemma 4 12B: fits with caveats (capacity estimate) (BF16, system RAM, CPU only); Gemma 4 E4B: fits with caveats (capacity estimate) (BF16, system RAM, CPU only); Gemma 4 E2B: fits with caveats (capacity estimate) (BF16, system RAM, CPU only); RigCheck capacity estimate, not a measured setup. The full checker takes your exact setup (context length, RAM, image input) and gives the command to start it.

Can a 16 GB graphics card run Qwen-Image-2.1 (image generation)?

In RigCheck's quick setup with 64 GB of system RAM, the checker has an answer on 3 of the 21 16 GB cards in the finder. On 3 of 21 cards: GeForce RTX 5080, GeForce RTX 5070 Ti, GeForce RTX 5060 Ti: Only as a planning estimate: the checker's uncalibrated memory budget, offload peak unverified: Above the all-components-resident planning budget; offload peak unverified; Uncalibrated allowances, not a minimum or a guarantee. Full checkpoint files are counted in host RAM even when mmap or sequential loading can use less. Actual peak depends on attention, prefix caches, VAE tiling, reference sizes and backend. No speed prediction. No answer on: GeForce RTX 4080 / 4080 Super, GeForce RTX 4070 Ti Super, GeForce RTX 4060 Ti, GeForce RTX 4090 Laptop, Tesla V100, Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT, Radeon RX 7600 XT, Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT, Radeon VII / Instinct MI50, Intel Arc A770. The full checker lets you change the task, resolution, references and offload (batch is fixed at one), and lists the files.

Can a 16 GB graphics card run MiniMax H3 (video + audio)?

In RigCheck's quick setup with 64 GB of system RAM, the checker has an answer on 7 of the 21 16 GB cards in the finder. On 7 of 21 cards: GeForce RTX 5080, GeForce RTX 5070 Ti, GeForce RTX 5060 Ti, GeForce RTX 4080 / 4080 Super, GeForce RTX 4070 Ti Super, GeForce RTX 4060 Ti: A published route or recipe matches, with caveats: Hardware memory matches a published video recipe; exact clip peak unverified; H3-Base FL2VA, source recipe canvases only. Disk-streaming routes assume an NVMe drive. Host-RAM and lab assumptions remain in the full checker; full 2K needs an API; Matching recipes: SGLang Recipe A (lab): one card, DiT, text encoder and VAE streamed from the checkpoint file on an NVMe drive (Recipe A), BF16; measured at 864×480 (at the native 768P the cookbook says to use video_vae=24 if the decode runs out of memory, not measured); lab run with VRAM capped at 12 GiB and a simulated 32 GB host: shows the memory fits, not your card's speed; a real 32 GB machine re-reads the weights from disk on every step. No answer on: GeForce RTX 4090 Laptop, Tesla V100, Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT, Radeon RX 7600 XT, Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT, Radeon VII / Instinct MI50, Intel Arc A770. The full checker lets you change the route or recipe settings and lists the files and the published recipe.

Can a 16 GB graphics card run Nemotron 3.5 (Bolt 74B-A9B, Super VL 120B-A12B)?

In RigCheck's quick setup with 64 GB of system RAM, the checker has an answer on 17 of the 21 16 GB cards in the finder. On 9 of 21 cards: Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT, Radeon RX 7600 XT, Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT: A published route or recipe matches, with caveats: Nemotron 3.5 Bolt 74B-A9B: a published route or recipe matches, with caveats (Matches Unsloth's short-context budget: about 32 GB of RAM or unified memory); Nemotron 3.5 Super VL 120B-A12B: a published route or recipe matches, with caveats (Matches Unsloth's 64 GB budget: UD-IQ1_M (46.1 GB) or UD-IQ2_XXS (51.1 GB)); Nemotron 3.5 Bolt 74B-A9B: We found no source covering AMD GPUs for this model, so only system RAM is counted here. Unsloth budgets around 32 GB of RAM or unified memory for short-context use (48–64 GB for more headroom) with its 22 GB IQ2_XXS GGUF in llama.cpp; longer contexts, images and video need more. This system RAM is 64 GB. That is Unsloth's planning estimate, not a measured run on this hardware. Nemotron 3.5 Super VL 120B-A12B: We found no source covering AMD or Intel GPUs for this model, so only system RAM is counted. Unsloth's planning estimate for a modest context, including the OS, runtime, KV cache and vision file; not a measured run. This setup counts 64 GB of RAM = 64 GB. Longer contexts and video need more. On 8 of 21 cards: GeForce RTX 5080, GeForce RTX 5070 Ti, GeForce RTX 5060 Ti, GeForce RTX 4080 / 4080 Super, GeForce RTX 4070 Ti Super, GeForce RTX 4060 Ti, GeForce RTX 4090 Laptop: A published route or recipe matches, with caveats: Nemotron 3.5 Bolt 74B-A9B: a published route or recipe matches, with caveats (Matches Unsloth's short-context budget: about 32 GB of RAM or unified memory); Nemotron 3.5 Super VL 120B-A12B: a published route or recipe matches, with caveats (Matches Unsloth's 80 GB budget: UD-IQ2_M, UD-Q2_K_XL or UD-IQ3_XXS (55.9–59.8 GB)); Nemotron 3.5 Bolt 74B-A9B: NVIDIA publishes no recipe for a 16 GB card; its discrete-GPU recipes start at 32 GB. Unsloth budgets around 32 GB of RAM or unified memory for short-context use (48–64 GB for more headroom) with its 22 GB IQ2_XXS GGUF in llama.cpp; longer contexts, images and video need more. This system RAM is 64 GB. That is Unsloth's planning estimate, not a measured run on this hardware. Nemotron 3.5 Super VL 120B-A12B: Unsloth's planning estimate for a modest context, including the OS, runtime, KV cache and vision file; not a measured run. This setup counts 16 GB of VRAM + 64 GB of RAM = 80 GB. Longer contexts and video need more. No answer on: Tesla V100, Radeon VII / Instinct MI50, Intel Arc A770. The full checker lets you change the route or recipe settings and lists the files and the published recipe.

Can a 16 GB graphics card run FLUX 3 Image Klein (int8, ComfyUI)?

In RigCheck's quick setup with 64 GB of system RAM, the checker has an answer on 8 of the 21 16 GB cards in the finder. On 8 of 21 cards: GeForce RTX 5080, GeForce RTX 5070 Ti, GeForce RTX 5060 Ti, GeForce RTX 4080 / 4080 Super, GeForce RTX 4070 Ti Super, GeForce RTX 4060 Ti, GeForce RTX 4090 Laptop: A published route or recipe matches, with caveats: Matches BFL's 8 GB guidance: the int8 files in ComfyUI; BFL: the int8 ComfyUI files run on 8 GB GPUs such as the RTX 3070. Image model, int8 text encoder and VAE together are 11.0 GB; by our reading, on a card smaller than that ComfyUI keeps them on the GPU one at a time and needs system RAM for the rest. BFL gives no RAM figure. The prompt upsampler BFL says not to skip is a 9B model (6.72 GB w4a8, 9.75 GB int8 or 18.8 GB BF16 in ComfyUI); BFL does not say whether its 8 GB figure includes it. When it does not fit with the image model, run it on its own first (BFL's README) or switch it off in the ComfyUI template, with weaker results. No answer on: Tesla V100, Radeon RX 9070 XT, Radeon RX 9070, Radeon RX 9060 XT, Radeon RX 7800 XT, Radeon RX 7600 XT, Radeon RX 6900 XT / 6950 XT, Radeon RX 6800 / 6800 XT, Radeon VII / Instinct MI50, Intel Arc A770. The full checker lets you change the route or recipe settings and lists the files and the published recipe.

Why do 16 GB cards get different answers?

Each checker judges the card itself, with its own support list and the routes its sources name, not only the memory size. On this page that happens for Qwen3.8-Flash-Next in the Strata engine, Qwen3.8-Flash-Next in llama.cpp, Mistral Medium 3.5 in llama.cpp, Qwen3.8-27B in llama.cpp, Gemma 4 (E2B, E4B, 12B, 26B-A4B, 31B) in llama.cpp, Qwen-Image-2.1 (image generation), MiniMax H3 (video + audio), Nemotron 3.5 (Bolt 74B-A9B, Super VL 120B-A12B), FLUX 3 Image Klein (int8, ComfyUI), GLM-5.3-Flash (Unsloth GGUF); the table lists each card's answer. Speeds also differ between cards with the same memory, so this page has no speed line: each card's own page and the full checkers show measured speeds, labeled.

How does system RAM change what a 16 GB card can run?

With 32 GB instead of 64 GB, the answer changes on some or all of these cards for 5 models: Qwen3.8-Flash-Next in the Strata engine, Mistral Medium 3.5 in llama.cpp, Qwen3.8-27B in llama.cpp, Gemma 4 (E2B, E4B, 12B, 26B-A4B, 31B) in llama.cpp, Nemotron 3.5 (Bolt 74B-A9B, Super VL 120B-A12B). With more than 64 GB, a checker gives a better answer for 1 model: Qwen3.8-Flash-Next in llama.cpp. See What system RAM changes for the cards and the answers.

Other memory sizes

8 GB · 12 GB · 24 GB · 32 GB · All VRAM sizes · Graphics cards · Macs · Any other machine

Sources