Qwen1.5 110B — Hardware Requirements & GPU Compatibility
Chat
Qwen1.5-110B is Alibaba's largest dense base model in the Qwen1.5 series, a pretrained language model that is not instruction-tuned, released as a beta preview ahead of Qwen2. Like the rest of the Qwen1.5 line it uses a SwiGLU Transformer with an improved multilingual tokenizer, and at 110B it is one of only two sizes in the series, alongside 32B, to include grouped-query attention. A matching aligned chat model was released alongside it. At 111 billion parameters it requires a multi-GPU workstation or server to run, even quantized. Context length is 32,768 tokens. It is released under Alibaba's custom Tongyi Qianwen license, which permits commercial use below 100 million monthly active users. It was published in April 2024; Qwen1.5 has since been superseded by the Qwen2, Qwen2.5, and Qwen3 series.
1.2K downloads
104 likesSpecifications
- Publisher
- Alibaba
- Family
- Qwen
- Parameters
- 111.2B
- Architecture
- Qwen2ForCausalLM
- Context Length
- 32,768 tokens
- Vocabulary Size
- 152,064
- Release Date
- 2024-04-25
- License
- Other
Get Started
HuggingFace
How Much VRAM Does Qwen1.5 110B Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| Q2_K | 3.40 | 48.2 GB | 58.3 GB | 47.26 GB | 2-bit quantization with K-quant improvements |
| Q3_K_S | 3.50 | 49.6 GB | 59.7 GB | 48.65 GB | 3-bit small quantization |
| Q3_K_Mest. | 3.90 | 55.2 GB | 65.3 GB | 54.21 GB | 3-bit medium quantization |
| Q4_K_Mest. | 4.80 | 67.7 GB | 77.8 GB | 66.73 GB | 4-bit medium quantization — most popular sweet spot |
| Q5_K_Mest. | 5.70 | 80.2 GB | 90.3 GB | 79.24 GB | 5-bit medium quantization — good quality/size tradeoff |
| Q6_Kest. | 6.60 | 92.7 GB | 102.8 GB | 91.75 GB | 6-bit quantization, very good quality |
| Q8_0est. | 8.00 | 112.2 GB | 122.3 GB | 111.21 GB | 8-bit quantization, near-lossless |
est.= calculated VRAM estimate; no published GGUF file found for that quantization yet. Other rows are verified against real community uploads.
Which GPUs Can Run Qwen1.5 110B?
Q4_K_M · 67.7 GBQwen1.5 110B (Q4_K_M) requires 67.7 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 89+ GB is recommended. Using the full 33K context window can add up to 10.1 GB, bringing total usage to 77.8 GB. No single GPU has enough memory — multi-GPU or cluster setups are needed.
Which Devices Can Run Qwen1.5 110B?
Q4_K_M · 67.7 GB19 devices with unified memory can run Qwen1.5 110B, including NVIDIA DGX H100, NVIDIA DGX A100 640GB, Mac Studio (M3 Ultra, 96GB).
Runs great
— Plenty of headroom
Decent
— Enough memory, may be tightWhere to Download Qwen1.5 110B
Community quantizations of this model — GGUF for llama.cpp, Ollama, and LM Studio, plus AWQ/MLX variants where available.
Benchmarks
Benchmark details →Related Models
Frequently Asked Questions
- How much VRAM does Qwen1.5 110B need?
Qwen1.5 110B requires 67.7 GB of VRAM at Q4_K_M, or 223.4 GB at BF16. Full 33K context adds up to 10.1 GB (77.8 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 111.2B × 4.8 bits ÷ 8 = 66.7 GB
KV Cache + Overhead ≈ 1 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 11.1 GB (at full 33K context)
VRAM usage by quantization
Q4_K_M67.7 GBQ4_K_M + full context77.8 GB- Can NVIDIA GeForce RTX 5090 run Qwen1.5 110B?
No — Qwen1.5 110B requires at least 46.9 GB at IQ3_XS, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
- What's the best quantization for Qwen1.5 110B?
For Qwen1.5 110B, Q4_K_M (67.7 GB) offers the best balance of quality and VRAM usage. Q5_K_M (80.2 GB) provides better quality if you have the VRAM. The smallest option is IQ3_XS at 46.9 GB.
VRAM requirement by quantization
IQ3_XS46.9 GBQ3_K_S49.6 GBQ3_K_M55.2 GBQ4_K_M ★67.7 GBQ6_K92.7 GBBF16223.4 GB★ Recommended — best balance of quality and VRAM usage.
- Can I run Qwen1.5 110B on a Mac?
Qwen1.5 110B requires at least 46.9 GB at IQ3_XS, which exceeds the unified memory of most consumer Macs. You would need a Mac Studio or Mac Pro with a high-memory configuration.
- Can I run Qwen1.5 110B locally?
Yes — Qwen1.5 110B can run locally on consumer hardware. At Q4_K_M quantization it needs 67.7 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Qwen1.5 110B?
At Q4_K_M, Qwen1.5 110B can reach ~71 tok/s on AMD Instinct MI350X. 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 ÷ 67.7 × 0.65 = ~77 tok/s
Estimated speed at Q4_K_M (67.7 GB)
~77 tok/s~77 tok/s~71 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Qwen1.5 110B?
At Q4_K_M, the download is about 66.73 GB. The full-precision BF16 version is 222.42 GB. The smallest option (IQ3_XS) is 45.87 GB.
- Which GPUs can run Qwen1.5 110B?
No single consumer GPU has enough VRAM to run Qwen1.5 110B at Q4_K_M (67.7 GB). Multi-GPU or professional hardware is required.
- Which devices can run Qwen1.5 110B?
19 devices with unified memory can run Qwen1.5 110B at Q4_K_M (67.7 GB), including ASUS Ascent GX10, Asus ROG Flow Z13 (2025, Ryzen AI Max+ 395, 128 GB), Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB), Framework Desktop (Ryzen AI Max+ 395, 128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.