Qwen3.5-9b-Sushi-Coder-RL needs about 7.2 GB of memory at Q4_K_M (a 5.6 GB file) with an 8K context window, so it fits on a 8 GB GPU. This repository has 3 quants, from 5.6 GB (Q4_K_M) to 9.5 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 4.3 GB.
Qwen3.5-9b-Sushi-Coder-RL: multimodal coding model
Qwen3.5-9b-Sushi-Coder-RL is a 8.95B parameter model merged with an RL adapter for coding tasks. It supports coding, reasoning, and vision in a single GGUF export. The model is licensed under Apache-2.0.
Good at
- Coding with LoRA fine‑tuning
- Reasoning on long contexts
- Multimodal vision tasks
Limits
- No vision support without the BF16-mmproj file
- Requires large VRAM for higher quantizations
- No mixture of experts
Best for
- Code generation in IDEs
- Automated code review tools
- Multimodal chat assistants
Who it’s for. Developers needing a compact coding assistant that can also process images. Researchers exploring RL‑fine‑tuned Qwen variants. Teams building multimodal chatbots with coding capabilities.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_K_M | 5.6 GB | 7.2 GB | Tight | Get file → | |
| Q6_K | 7.4 GB | 8.9 GB | Too big | Get file → | |
| Q8_0 | 9.5 GB | 11.1 GB | Too big | Get file → |
For 8 GB VRAM, use the Q4_K_M quant (file size 5.63 GB, estimated 7.2 GB at 8 k context). If 12 GB VRAM is available, Q6_K (7.36 GB, 8.9 GB) is a good balance. For maximum accuracy and 16 GB VRAM, Q8_0 (9.53 GB, 11.1 GB) is recommended.
What GPU do you need for Qwen3.5-9b-Sushi-Coder-RL?
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: no quant fits at an 8K context.
- 8 GB VRAM: the best quant that fits is Q4_K_M (5.6 GB file, about 7.2 GB in use).
- 12 GB VRAM: the best quant that fits is Q8_0 (9.5 GB file, about 11.1 GB in use).
Which Qwen3.5-9b-Sushi-Coder-RL quant should you choose?
Q4_K_M is the usual starting point for Qwen3.5-9b-Sushi-Coder-RL: it balances file size and quality well. Here it is a 5.6 GB file.
The Q8_0 file is 9.5 GB, about 1.7 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 Qwen3.5-9b-Sushi-Coder-RL
How much VRAM does Qwen3.5-9b-Sushi-Coder-RL need?
About 7.2 GB at Q4_K_M with an 8K context window. That is the 5.6 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 Qwen3.5-9b-Sushi-Coder-RL on an 8 GB GPU?
Yes, with Q4_K_M: the 5.6 GB file needs about 7.2 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Qwen3.5-9b-Sushi-Coder-RL GGUF?
It depends on the quant: Q4_K_M is 5.6 GB, Q6_K is 7.4 GB, Q8_0 is 9.5 GB.
What context length does Qwen3.5-9b-Sushi-Coder-RL support?
The GGUF header lists a trained context length of 262,144 tokens (256K). 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 Qwen3.5-9b-Sushi-Coder-RL GGUF?
The files are on Hugging Face in the bigatuna/Qwen3.5-9b-Sushi-Coder-RL-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does Qwen3.5-9b-Sushi-Coder-RL 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 Qwen3.5-9b-Sushi-Coder-RL on 8 GB VRAM?
Yes, the Q4_K_M quant uses about 7.2 GB at 8 k context, fitting within 8 GB VRAM.
Do I need the BF16-mmproj file for vision?
Yes, the text GGUF must be used together with the BF16-mmproj file for multimodal vision support.
Is the model licensed for commercial use?
The model is released under Apache-2.0, which permits commercial use with attribution.