Datasheet for local AI49 models98 GPUsData read 2026-09-25
No adsNo tracking
Check my GPU →

Can the Mac 24 GB run LTX-2.5 (22B)?

Video model · 22B24 GB unified · ~19.6 GB for the GPU · 100–307 GB/s · Apple SiliconData 2026-09-25
Tight
Only just.

Download Q3_K_M (11.5 GB). With the model's working memory it needs about 16.3 GB, leaving 3.3 GB spare on 19.6 GB. Quality: noticeable loss of detail. For better quality, Q4_K_M (15.1 GB) also runs, with about 0.3 GB spilling into system RAM — a little slower, still practical.

Best fileQ3_K_M
File size11.5 GB
GPU memory needed~16.3 GB
Shared memory46 / 24 GB
Memory map · Mac 24 GB16.3 GB / 19.6 GB
010 GB20 GB
Weights Q3_K_M · 11.5 GBWorking memory · 4.0 GBReserve · 0.8 GBFree · 3.3 GB
calculated from real file sizes plus working memory. Not yet measured on this setup. How this works.

01What to download

PartFileFolderSize
ModelLTX-2.5-Distilled-Q3_K_M.gguf
Q3_K_M · realrebelai/LTX-2.5_GGUFs
models/unet11.5 GBDownload →
Text encodergemma4-12b-with-proj-ltx-2.5-bf16.safetensors
Gemma 4 12B + LTX projection BF16 · Lightricks/LTX-2.5
models/text_encoders26.3 GBDownload →
VAEltx-2.5-video-vae-bf16.safetensors
LTX-2.5 video VAE (DiffVAE) BF16 · Lightricks/LTX-2.5
models/vae1.5 GBDownload →
Also neededltx-2.5-audio-vae-bf16.safetensors
LTX-2.5 audio VAE + vocoder BF16 · Lightricks/LTX-2.5
models/vae0.4 GBDownload →
Total download · keep about the same free on disk39.6 GB

Only for some workflows: LTX-2.5 latent spatial upscaler x2 (1.0 GB, official two-stage template); Gemma 4 E2B INT8 (prompt enhancer, optional) (5.2 GB, only if prompt enhancer enabled). On a Mac, FP8 text encoders do not load, so a 16-bit one is listed. Sizes read from Hugging Face (2026-09-25). “Download” links start the file directly; the file name links are the same files the official ComfyUI workflows use.

Shared memory: about 45.6 GB of your 24 GB. On this machine the model, the text encoder, the VAE and the operating system all use the same memory. That is about 21.6 GB more than there is. It still runs: while the model works, the system compresses or swaps the idle text encoder, so loading and changing the prompt get slower. A smaller file or text encoder avoids that. calculated

02Every LTX-2.5 file on 19.6 GB

FileSizeNeededOn this cardQualityDownload
16-bit
SAFETENSORS · Lightricks
42.0 GB46.8 GB27.2 GB too bigthe original weightsHugging Face →
Q8_0
GGUF · realrebelai
23.6 GB28.4 GB8.8 GB too bigpractically identical to the originalHugging Face →
Q6_K
GGUF · realrebelai
18.7 GB23.5 GB3.9 GB too bigvery close to the originalHugging Face →
Q5_K_M
GGUF · realrebelai
16.8 GB21.6 GB2.0 GB too bigclose; small differences in fine detailHugging Face →
Q4_K_M
GGUF · realrebelai
15.1 GB19.9 GBspills 0.3 GBgood; some loss in fine detail and textHugging Face →
Q3_K_M ←
GGUF · realrebelai
11.5 GB16.3 GBfits · 3.3 GB sparenoticeable loss of detailHugging Face →
Q2_K
GGUF · realrebelai
8.8 GB13.6 GBfits · 6.0 GB spareheavy loss; a last resortHugging Face →

Sizes from the Hugging Face file listing, read 2026-09-25. “Needed” = file + 4 GB working memory for this model + 0.8 GB kept free for the system.

03The text encoder

Gemma 4 12B (custom LTX-2.5 build): 26.3 GB as 16-bit, 15.4 GB as INT8. ComfyUI encodes the prompt first and can push the encoder out of VRAM before sampling, so it does not have to fit together with the model. On 19.6 GB the INT8 encoder fits on its own, so prompt encoding stays fast.

04About this GPU

A Mac with 24 GB shares that memory between the CPU and the GPU. On current macOS the GPU may use about 19.6 GB of it by default (older macOS versions: about 17.2 GB); that is the figure used here. ComfyUI can go past it, but macOS then starts compressing and swapping memory and everything slows down. ComfyUI runs on Apple GPUs through PyTorch's MPS backend. 16-bit and GGUF files work; FP8 and INT8 files do not save memory on a Mac, so they are skipped. Speed is the catch: even the fastest Macs are several times slower per image than a desktop RTX card. Chips sold with 24 GB: M2, M3, M4, M4 Pro, M5, M5 Pro, M6 (memory bandwidth 100–307 GB/s — the higher, the faster). Everything about Macs and local AI →

Mac 24 GB: 24 GB unified · ~19.6 GB for the GPU · 100–307 GB/s · Apple Silicon. Everything that runs on the Mac 24 GB →

05What more VRAM would change

With 32 GB you could run LTX-2.5 at 8-bit or better (Q8_0, 23.6 GB) with no offloading: for example on the RTX 5090 32 GB.

06Measured and reported results

Nobody has sent measured numbers for this pair yet. If you run LTX-2.5 on a Mac 24 GB, send your time per image and peak VRAM and it will appear here, credited.

07Questions

How much VRAM does LTX-2.5 need?

Around 13.6 GB with the smallest file (Q2_K) and 28.4 GB with an 8-bit file (Q8_0), counting working memory and a small system reserve. The full 16-bit file needs about 46.8 GB.

Which LTX-2.5 file should I download for the Mac 24 GB?

Q3_K_M (11.5 GB) from realrebelai/LTX-2.5_GGUFs. Load it with Unet Loader (GGUF) from the ComfyUI-GGUF node pack.

Is FP8 faster than GGUF on the Mac 24 GB?

No. This GPU has no FP8 compute, so ComfyUI converts FP8 weights back before the maths. FP8 only saves memory here; Q8_0 GGUF is the closer-to-original 8-bit pick.

How much do I need to download for LTX-2.5 on the Mac 24 GB?

About 39.6 GB for the model file, text encoder and VAE listed on this page, and the same again free on disk. With a 19.6 GB card, 48 GB of system RAM or more is recommended.