RTX 4060 Ti 8 GB for local AI
Which image and video models run on the RTX 4060 Ti 8 GB, which file to download for each, and how much VRAM they need.
VRAM8 GB
MemoryGDDR6
Bus128-bit
Bandwidth288 GB/s
ArchitectureAda Lovelace
Launched2023-05
Launch price$399
Runs well8 of 49
Specs: www.nvidia.com · launch: en.wikipedia.org
01What runs on it
| Model | Verdict | Best file | Size | Needed |
|---|---|---|---|---|
| image models | ||||
| FLUX.1 [dev] | Tight | Q3_K_S | 5.2 GB | 7.5 GB |
| FLUX.1 [schnell] | Tight | Q3_K_S | 5.2 GB | 7.5 GB |
| FLUX.1 Kontext [dev] | Tight | Q3_K_M | 5.4 GB | 8.0 GB |
| FLUX.1 Krea [dev] | Tight | Q3_K_M | 5.4 GB | 7.7 GB |
| FLUX.1 Fill [dev] | Tight | Q3_K_S | 5.2 GB | 7.8 GB |
| FLUX.2 [dev] | Not practical | Q2_K | 12.9 GB | 16.2 GB |
| FLUX.2 [klein] 9B | Tight | Q3_K_M | 4.8 GB | 7.1 GB |
| FLUX.2 [klein] 4B | Runs well | FP8 | 4.1 GB | 5.9 GB |
| Krea 2 (Turbo) | Tight | Q2_K | 4.9 GB | 7.5 GB |
| Qwen-Image | Offload only | Q2_K | 7.1 GB | 9.9 GB |
| Qwen-Image-Edit (2511) | Offload only | Q4_K_M | 13.2 GB | 16.2 GB |
| Qwen-Image 2.1 | Runs | Q5_K_M | 5.0 GB | 7.3 GB |
| Z-Image Turbo | Runs | Q6_K | 5.9 GB | 7.9 GB |
| Z-Image (base) | Runs | Q5_K_M | 5.6 GB | 7.6 GB |
| Ideogram 4 | Offload only | Q4_1 | 6.2 GB | 14.7 GB |
| Boogu-Image (Turbo) | Offload only | Q4_1 | 7.4 GB | 10.0 GB |
| ERNIE-Image (Turbo) | Runs | Q4_K_M | 5.0 GB | 7.3 GB |
| HiDream-O1-Image | Offload only | FP8 | 8.1 GB | 10.9 GB |
| Mage-Flow (Microsoft) | Runs well | INT8 | 4.2 GB | 6.0 GB |
| Lumina Image 2.0 | Runs well | 16-bit | 5.2 GB | 7.0 GB |
| HiDream-I1 (Full) | Offload only | Q2_K | 6.6 GB | 9.4 GB |
| HiDream-I1 (Dev) | Offload only | Q2_K | 6.6 GB | 9.4 GB |
| Stable Diffusion 3.5 Large | Runs | Q4_1 | 5.3 GB | 7.6 GB |
| Stable Diffusion 3.5 Medium | Runs well | 16-bit | 5.1 GB | 6.9 GB |
| Chroma1-HD | Runs | Q4_K_M | 5.6 GB | 7.9 GB |
| SDXL 1.0 | Runs well | 16-bit | 6.9 GB | 7.1 GB |
| Illustrious XL / Pony (SDXL anime) | Runs well | 16-bit | 6.9 GB | 7.1 GB |
| Stable Diffusion 1.5 | Runs well | 16-bit | 2.1 GB | 3.3 GB |
| HunyuanImage 2.1 | Offload only | Q4_K_M | 11.3 GB | 14.6 GB |
| video models | ||||
| Wan 2.1 T2V 14B | Offload only | Q4_K_M | 10.1 GB | 14.4 GB |
| Wan 2.1 T2V 1.3B | Runs well | 16-bit | 2.8 GB | 5.6 GB |
| Wan 2.1 I2V 14B 480P | Offload only | Q4_K_M | 11.3 GB | 15.6 GB |
| Wan 2.1 I2V 14B 720P | Offload only | Q4_K_M | 11.3 GB | 18.1 GB |
| Wan 2.1 VACE 14B | Offload only | Q4_K_M | 11.6 GB | 16.4 GB |
| Wan 2.2 T2V A14B | Offload only | Q2_K | 5.3 GB | 9.6 GB |
| Wan 2.2 I2V A14B | Offload only | Q2_K | 5.3 GB | 9.6 GB |
| Wan 2.2 TI2V 5B | Runs | Q5_K_M | 3.8 GB | 7.6 GB |
| Wan 2.2 Animate 14B | Offload only | Q4_K_M | 11.5 GB | 16.8 GB |
| Wan Animate 2 (14B) | Offload only | Q4_K_M | 11.3 GB | 16.6 GB |
| Wan 2.2 S2V 14B | Offload only | Q4_K_M | 13.9 GB | 18.7 GB |
| SCAIL-2 (character animation) | Offload only | Q4_K_M | 11.5 GB | 16.8 GB |
| HunyuanVideo (13B, original) | Offload only | Q4_K_M | 7.9 GB | 12.2 GB |
| LTX-Video 13B (0.9.8) | Offload only | Q2_K | 4.7 GB | 9.0 GB |
| HunyuanVideo 1.5 | Offload only | Q4_K_M | 5.1 GB | 9.4 GB |
| LTX-2 (19B) | Offload only | Q4_K_M | 12.8 GB | 17.6 GB |
| LTX-2.3 (22B) | Offload only | Q4_K_M | 14.3 GB | 19.1 GB |
| LTX-2.5 (22B) | Offload only | Q4_K_M | 15.1 GB | 19.9 GB |
| MiniMax H3 (33B) | Not practical | Q3_K_M | 15.6 GB | 21.4 GB |
| MiniMax H3 Pruned | Offload only | Q4_K_M | 11.6 GB | 17.4 GB |
Calculated from real file sizes plus working memory. How the numbers work.
02Good to know
The RTX 4060 Ti 8 GB is an Ada Lovelace GPU with hardware FP8, so ComfyUI can compute Comfy-Org's FP8 files natively: small and fast. (Plain FP8 files use FP8 maths with the --fast fp8_matrix_mult option.)
03Measured and reported results
| Label | Model | Setup | Result | Peak VRAM | Date | Source |
|---|---|---|---|---|---|---|
| reported | FLUX.1 dev | FP8 “Flux-Dev(FP8) … 2回目は72秒から51秒とあまり効果が感じられませんでした。” ComfyUI; same PC before/after swapping RTX 4060 Ti 8GB -> 16GB; 2nd-run times; resolution/steps not stated; 72 s = 8GB card | 72 s / image | — | 2025-01-03 | note.com → |
| reported | FLUX.1 Krea | Flux.1 Krea Dev CLIP+VAE FP8 (~20GB checkpoint) · 768x1024 · 20 steps “すべて768x1024または1024x768、20ステップ、seed固定。… 合計（2回目以降） | 約40秒” ComfyUI with async weight offloading to system RAM; sampling ~35 s; first run ~130 s; approximate ('約'); 768x1024 or 1024x768 | 40 s / image | — | 2026-05-06 | hide10.com → |
| reported | FLUX.1 schnell | FP8 “Flux-Schnell(FP8) … 2回目は24秒から11秒です。” ComfyUI; same PC before/after swapping RTX 4060 Ti 8GB -> 16GB; 2nd-run times; resolution/steps not stated; 24 s = 8GB card | 24 s / image | — | 2025-01-03 | note.com → |
Reported results are other people's numbers, copied as published, with a link. Settings, drivers and ComfyUI versions differ, so compare them with care. Send yours.