Llm Jp 3.1 13B Instruct4 — Hardware Requirements & GPU Compatibility
Chat
Llm Jp 3.1 13B Instruct4 is a 13.7B-parameter open language model from llm-jp. It supports a context window of up to 4,096 tokens. At Q4_K_M it needs about 10.20 GB of VRAM — see which GPUs and Macs can run it below.
Runs well on a 24 GB GPU
Runs from 12 GB (Q4_K_M · 4K)Best on 24 GB (Q4_K_M · ~64 tok/s · 4K)Macs from 16 GB (Q4_K_M · ~14 tok/s)
Apache 2.0
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run Llm Jp 3.1 13B Instruct4
One recommended way. Every other method, OS and quant is in the step-by-step guide.
Recommended
vllm serve llm-jp/llm-jp-3.1-13b-instruct4 --max-model-len 4096vLLM repo with its own quantisation
Download: Weights on Hugging Face
Other ways
each opens step-by-stepRun in cloud
Fits on RTX A4000 (16 GB) (4 GB headroom) · Q4_K_M
- Generation speed
- ~29 tok/s
- generation speed
- Cost per 1M output tokens
- $1.07
- per 1M output tokens
How Much VRAM Does Llm Jp 3.1 13B Instruct4 Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 7.8 GB | — | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 4K context · ~84 tok/s | |
| 8.7 GB | — | 3-bit medium — 3-bit medium quantization | Fits: 4K context · ~76 tok/s | |
| 10.2 GB | — | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 4K context · ~64 tok/s | |
| 11.7 GB | — | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 4K context · ~56 tok/s | |
| 13.3 GB | — | 6-bit, very good — 6-bit quantization, very good quality | Fits: 4K context · ~49 tok/s | |
| 15.7 GB | — | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 4K context · ~42 tok/s | |
| 29.4 GB | — | 16-bit · bfloat — Brain floating point 16 — preferred for training | Does not fit: needs 29.4 GB |
Selected Q4_K_M
4-bit medium quantization11.9GBat 4K context
- Weights
- 8.2 GB
- KV cache
- 3.4 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 12.1 GB free · ~64 tok/s
≥50 tok/s needs ≥784 GB/s · RTX 5090 ~114, RTX 3090 Ti ~64
Which Hardware Can Run Llm Jp 3.1 13B Instruct4?
Q4_K_M · 10.2 GBLlm Jp 3.1 13B Instruct4 (Q4_K_M) requires 10.2 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 14+ GB is recommended. Using the full 4K context window can add up to 1.7 GB, bringing total usage to 11.9 GB. 38 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti, NVIDIA GeForce RTX 3080 Ti. 150 devices with unified memory can run Llm Jp 3.1 13B Instruct4, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), Mac Mini M2 Pro (16 GB).
12 GB
12 GB tier →fits · 0.1 GB free · up to 4K ctx
10 GPUs
RTX 3080 Ti~58Simulate output on RTX 3080 TiRTX 5070~43Simulate output on RTX 5070RTX 4070~32Simulate output on RTX 407020 GB
fits · 8.1 GB free · up to 4K ctx
1 GPU
RX 7900 XT~47Simulate output on RX 7900 XT32–40 GB
fits · 20.1 GB free · up to 4K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~114Simulate output on RTX 5090M5 Max36 GB~32Simulate output on Mac Studio M5 Max (36 GB)M532 GB~11Simulate output on MacBook Air 13" M5 (32 GB)Orin~13Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~9Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Similar models
6| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| Phi 4 ReasoningMicrosoft | — | 12.5 GB | ~69 | same size |
| 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 |
Specifications
- Publisher
- llm-jp
- Parameters
- 13.7B
- Architecture
- LlamaForCausalLM
- Context Length
- 4,096 tokens
- Vocabulary Size
- 99,584
- Release Date
- 2025-05-23
- License
- Apache 2.0
Frequently Asked Questions
How much VRAM does Llm Jp 3.1 13B Instruct4 need?
Llm Jp 3.1 13B Instruct4 requires 10.2 GB of VRAM at Q4_K_M, or 29.4 GB at BF16. Full 4K context adds up to 1.7 GB (11.9 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 13.7B × 4.8 bits ÷ 8 = 8.2 GB
KV Cache + Overhead ≈ 2 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.
KV Cache + Overhead ≈ 3.7 GB (at full 4K context)
VRAM usage by quantization
Can NVIDIA GeForce RTX 4090 run Llm Jp 3.1 13B Instruct4?
Yes, at Q8_0 (15.7 GB) or lower. Higher quantizations like BF16 (29.4 GB) exceed the NVIDIA GeForce RTX 4090's 24 GB.
What's the best quantization for Llm Jp 3.1 13B Instruct4?
For Llm Jp 3.1 13B Instruct4, Q4_K_M (10.2 GB) offers the best balance of quality and VRAM usage. Q5_K_M (11.7 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 7.8 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run Llm Jp 3.1 13B Instruct4 on a Mac?
Yes — Mac Mini M2 Pro (16 GB) and 133 other Macs can run Llm Jp 3.1 13B Instruct4. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 10.2 GB of usable unified memory (RAM minus macOS overhead).
Can I run Llm Jp 3.1 13B Instruct4 locally?
Yes — Llm Jp 3.1 13B Instruct4 can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Llm Jp 3.1 13B Instruct4?
At Q4_K_M, Llm Jp 3.1 13B Instruct4 can reach ~471 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~64 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 ÷ 10.2 × 0.65 = ~510 tok/s
Estimated speed at Q4_K_M (10.2 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of Llm Jp 3.1 13B Instruct4?
At Q4_K_M, the download is about 8.22 GB. The full-precision BF16 version is 27.42 GB. The smallest option (Q2_K) is 5.83 GB.
Which GPUs can run Llm Jp 3.1 13B Instruct4?
38 consumer GPUs can run Llm Jp 3.1 13B Instruct4 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. 26 GPUs have plenty of headroom for comfortable inference.
Which devices can run Llm Jp 3.1 13B Instruct4?
150 devices with unified memory can run Llm Jp 3.1 13B Instruct4 at Q4_K_M (10.2 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, 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.