ace step 1.5 gguf merge models needs about 4.6 GB of memory at Q4_K_M (a 3.0 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 20 quants, from 3.0 GB (Q4_K_M) to 10.0 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 4.3 GB.
ACE-Step 1.5 XL DiT models merged with task arithmetic
ACE-Step 1.5 XL DiT models are merged using task arithmetic to blend two fine‑tuned checkpoints. The models retain the XL architecture and support 32,768 token context. They are provided in multiple quantization levels for compatibility with llama.cpp and other C++ inference engines.
Good at
- Merges strengths of two fine‑tuned models
- Supports large 32k token context
- Available in multiple quant levels
Limits
- No vision support, only DiT
- Requires companion LM, text encoder, and VAE
- Higher step counts needed for quality
Best for
- Fine‑tuning with blended task vectors
- Generating high‑quality diffusion outputs
- Experimenting with interpolation ratios
Who it’s for. Developers and researchers who need a DiT model that combines the strengths of two fine‑tuned ACE-Step checkpoints. The model is suitable for projects requiring large context and flexible quantization.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_K_M | 3.0 GB | 4.6 GB | Fits | Get file → | |
| Q4_K_M | 3.0 GB | 4.6 GB | Fits | Get file → | |
| Q4_K_M | 3.0 GB | 4.6 GB | Fits | Get file → | |
| Q4_K_M | 3.0 GB | 4.6 GB | Fits | Get file → | |
| Q5_K_M | 3.5 GB | 5.1 GB | Fits | Get file → | |
| Q5_K_M | 3.5 GB | 5.1 GB | Fits | Get file → | |
| Q5_K_M | 3.5 GB | 5.1 GB | Fits | Get file → | |
| Q5_K_M | 3.5 GB | 5.1 GB | Fits | Get file → | |
| Q6_K | 4.1 GB | 5.7 GB | Fits | Get file → | |
| Q6_K | 4.1 GB | 5.7 GB | Fits | Get file → | |
| Q6_K | 4.1 GB | 5.7 GB | Fits | Get file → | |
| Q6_K | 4.1 GB | 5.7 GB | Fits | Get file → | |
| Q8_0 | 5.3 GB | 6.9 GB | Fits | Get file → | |
| Q8_0 | 5.3 GB | 6.9 GB | Fits | Get file → | |
| Q8_0 | 5.3 GB | 6.9 GB | Fits | Get file → | |
| Q8_0 | 5.3 GB | 6.9 GB | Fits | Get file → | |
| BF16 | 10.0 GB | 11.6 GB | Too big | Get file → | |
| BF16 | 10.0 GB | 11.6 GB | Too big | Get file → | |
| BF16 | 10.0 GB | 11.6 GB | Too big | Get file → | |
| BF16 | 10.0 GB | 11.6 GB | Too big | Get file → |
For 8 GB VRAM, use Q4_K_M (≈4.6 GB memory at 8k context). For 12 GB VRAM, Q5_K_M (≈5.1 GB) or Q6_K (≈5.7 GB) are suitable. For 16 GB VRAM, Q8_0 (≈6.9 GB) or BF16 (≈11.6 GB) can be used, depending on precision needs.
What GPU do you need for ace step 1.5 gguf merge models?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: the best quant that fits is Q6_K (4.1 GB file, about 5.7 GB in use).
- 8 GB VRAM: the best quant that fits is Q8_0 (5.3 GB file, about 6.9 GB in use).
- 12 GB VRAM: the best quant that fits is BF16 (10.0 GB file, about 11.6 GB in use).
Which ace step 1.5 gguf merge models quant should you choose?
Q4_K_M is the usual starting point for ace step 1.5 gguf merge models: it balances file size and quality well. Here it is a 3.0 GB file.
The Q8_0 file is 5.3 GB, about 1.8 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about ace step 1.5 gguf merge models
How much VRAM does ace step 1.5 gguf merge models need?
About 4.6 GB at Q4_K_M with an 8K context window. That is the 3.0 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run ace step 1.5 gguf merge models on an 8 GB GPU?
Yes, with Q8_0: the 5.3 GB file needs about 6.9 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of ace step 1.5 gguf merge models GGUF?
It depends on the quant: Q4_K_M is 3.0 GB, Q4_K_M is 3.0 GB, Q4_K_M is 3.0 GB, Q4_K_M is 3.0 GB, Q5_K_M is 3.5 GB, Q5_K_M is 3.5 GB and 14 more.
What context length does ace step 1.5 gguf merge models support?
The GGUF header lists a trained context length of 32,768 tokens (32K). The KV cache grows with the context you use: about 1.1 GB at 8K and 4.3 GB at 32K.
Where can I download ace step 1.5 gguf merge models GGUF?
The files are on Hugging Face in the scragnog/ace-step-1.5-gguf-merge-models repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does ace step 1.5 gguf merge models use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run Q4_K_M on 8 GB VRAM?
Yes, Q4_K_M uses about 4.6 GB of VRAM at 8k context, fitting within an 8 GB budget.
Do these models support vision tasks?
No, they are DiT-only and require companion LM, text encoder, and VAE models.
What is the context length of these models?
They support a context length of 32,768 tokens.