Qwen3.8 13.5B — Hardware Requirements & GPU Compatibility
Chat
Qwen3.8 13.5B is a 13.5B-parameter open language model from kusuriya-labs in the Qwen 3.8 family. At Q4_K_M it needs about 8.91 GB of VRAM — see which GPUs and Macs can run it below.
Runs great on a 24 GB GPU
Runs from 12 GB (Q4_K_M · 16K)Best on 24 GB (Q4_K_M · ~74 tok/s · 32K)Macs from 16 GB (Q4_K_M · ~16 tok/s)
Apache 2.0
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run Qwen3.8 13.5B
One recommended way. Every other method, OS and quant is in the step-by-step guide.Each way opens a step-by-step guide for your OS and quant.
Recommended
vllm serve kusuriya-labs/Qwen3.8-13.5B --max-model-len 16384vLLM repo with its own quantisation
Download: Weights on Hugging Face
Ways to runOther ways
each opens step-by-stepRun in cloud
Runs on RTX 3090 (24 GB) (12 GB headroom) · Q4_K_M
- Generation speed
- ~68 tok/s
- generation speed
- Cost per 1M output tokens
- $0.58
- per 1M output tokens
How Much VRAM Does Qwen3.8 13.5B Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 6.3 GB | 12.7 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 32K context · ~104 tok/s | |
| 7.2 GB | 13.6 GB | 3-bit medium — 3-bit medium quantization | Fits: 32K context · ~91 tok/s | |
| 7.4 GB | 13.8 GB | 4-bit · legacy — 4-bit legacy quantization | Fits: 32K context · ~88 tok/s | |
| 8.9 GB | 15.3 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 32K context · ~74 tok/s | |
| 10.6 GB | 17 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 32K context · ~62 tok/s | |
| 12.3 GB | 18.6 GB | 6-bit, very good — 6-bit quantization, very good quality | Fits: 32K context · ~54 tok/s | |
| 14.8 GB | 21.2 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 32K context · ~44 tok/s | |
| 29.7 GB | 36.1 GB | 16-bit · bfloat — Brain floating point 16 — preferred for training | Does not fit: needs 32.7 GB |
Selected Q4_K_M
4-bit medium quantization11.9GBat 16K context
- Weights
- 8.2 GB
- KV cache
- 3.4 GB
- Overhead
- 0.3 GB
Which Hardware Can Run Qwen3.8 13.5B?
Q4_K_M · 8.9 GBQwen3.8 13.5B (Q4_K_M) requires 8.9 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 12+ GB is recommended. 51 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti, NVIDIA GeForce RTX 3080 Ti. 171 devices with unified memory can run Qwen3.8 13.5B, including Mac Studio M5 Ultra (256 GB), Mac Studio M5 Ultra (512 GB), Mac Mini M2 Pro (16 GB).
12 GB
12 GB tier →fits · 0.1 GB free · up to 16K ctx
16 GPUs
RTX 3080 Ti~67Simulate output on RTX 3080 TiRTX 3080 12GB~67Simulate output on RTX 3080 12GBRTX 5070~49Simulate output on RTX 507020 GB
fits · 8.1 GB free · up to 32K ctx
1 GPU
RX 7900 XT~54Simulate output on RX 7900 XT32–40 GB
fits · 20.1 GB free · up to 32K ctx
1 GPU · 3 Mac classes · 9 device families
RTX 5090~131Simulate output on RTX 5090M5 Max*36 GB~36Simulate output on Mac Studio M5 Max (36 GB)M632 GB~13Simulate output on iMac 24" M6 (32 GB)RTX Spark N1X~22Simulate output on NVIDIA RTX Spark N1X Laptop (32 GB)Orin~15Simulate output on NVIDIA Jetson AGX Orin 32GB
By VRAM tier · typical GPUs and Macs at Q4_K_M
Alternatives to Qwen3.8 13.5B
Qwen 3.8 family · 13 models →Similar models
6| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| Phi 3 Medium 128k InstructMicrosoft | — | 12 GB | ~72 | same size |
| Qwen1.5 14BAlibaba | — | 22.2 GB | ~63 | same size |
| Baichuan2 13B Basebaichuan-inc | — | 11.5 GB | ~76 | same size |
| Llama 2 13B Chat HFMeta | — | 11.5 GB | ~76 | same size |
| Vicuna 13B V1.3LMSYS | — | 8.6 GB | ~76 | same size |
| Vllm Translategemma 12B ITchbae624 | — | 9.8 GB | ~73 | same size |
Specifications
- Publisher
- kusuriya-labs
- Family
- Qwen 3.8
- Parameters
- 13.5B
- Release Date
- 2026-11-09
- License
- Apache 2.0
Frequently Asked Questions
How much VRAM does Qwen3.8 13.5B need?
Qwen3.8 13.5B requires 8.9 GB of VRAM at Q4_K_M, or 29.7 GB at BF16.
VRAM = Weights + KV Cache + Overhead
Weights = 13.5B × 4.8 bits ÷ 8 = 8.1 GB
KV Cache + Overhead ≈ 0.8 GB (at 2K context + ~0.3 GB framework)
Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.
VRAM usage by quantization
Can NVIDIA GeForce RTX 4090 run Qwen3.8 13.5B?
Yes, at Q8_0 (14.8 GB) or lower. Higher quantizations like BF16 (29.7 GB) exceed the NVIDIA GeForce RTX 4090's 24 GB.
What's the best quantization for Qwen3.8 13.5B?
For Qwen3.8 13.5B, Q4_K_M (8.9 GB) offers the best balance of quality and VRAM usage. Q5_K_M (10.6 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 6.3 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run Qwen3.8 13.5B on a Mac?
Yes — Mac Mini M2 Pro (16 GB) and 145 other Macs can run Qwen3.8 13.5B. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 8.9 GB of usable unified memory (RAM minus macOS overhead).
Can I run Qwen3.8 13.5B locally?
Yes — Qwen3.8 13.5B can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Qwen3.8 13.5B?
At Q4_K_M, Qwen3.8 13.5B can reach ~539 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~74 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: NVIDIA B200 → 8000 ÷ 8.9 × 0.65 = ~584 tok/s
Estimated speed at Q4_K_M (8.9 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of Qwen3.8 13.5B?
At Q4_K_M, the download is about 8.10 GB. The full-precision BF16 version is 27.00 GB. The smallest option (Q2_K) is 5.74 GB.
Which GPUs can run Qwen3.8 13.5B?
51 consumer GPUs can run Qwen3.8 13.5B at Q4_K_M. Top options include AMD Radeon RX 6800, AMD Radeon RX 6800 XT, AMD Radeon RX 6900 XT, AMD Radeon RX 6700 XT. 30 GPUs have plenty of headroom for comfortable inference.
Which devices can run Qwen3.8 13.5B?
171 devices with unified memory can run Qwen3.8 13.5B at Q4_K_M (8.9 GB), including AMD Ryzen 7 8845HS (Hawk Point) Laptop, AMD Ryzen AI 7 350 (Krackan Point) Laptop, AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10. Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.