Hermes 3 Llama 3.2 3B — Hardware Requirements & GPU Compatibility
ChatRoleplay
Hermes 3 Llama 3.2 3B is a 3-billion parameter instruction-tuned model by Nous Research, fine-tuned from Meta's Llama 3.2 3B base. It applies the Hermes training methodology to a compact model, targeting strong instruction following and conversational quality at minimal hardware cost. Despite its small size, this model benefits from the Hermes fine-tuning approach that emphasizes system prompt adherence and structured output. It can run on GPUs with as little as 4GB of VRAM when quantized, making it suitable for lightweight local deployments and resource-constrained environments.
Runs great on a 24 GB GPU
Runs from 8 GB (Q4_K_M · 32K)Best on 24 GB (Q4_K_M · ~281 tok/s · 128K)Macs from 8 GB (Q4_K_M · ~31 tok/s)
Llama 3 Community
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run Hermes 3 Llama 3.2 3B
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
ollama run hermes3:3bQ4_K_M fits fully on the RTX 4090 · 20.0 GB free · 16K context
Download: Hermes-3-Llama-3.2-3B-Q4_K_M.gguf2.02 GBbartowski
Ways to runOther ways
each opens step-by-stepRun in cloud
Your own private copy · Q4_K_M · 24 GB GPU
- The easiest way to try this model
- Running in a few minutes, easy setup
- Chat right here or connect any app
$5+ bonus for new users
Runs on RTX 3060 12GB (8 GB headroom) · Q4_K_M
- Generation speed
- ~100 tok/s
- generation speed
- Cost per 1M output tokens
- $0.16
- per 1M output tokens
How Much VRAM Does Hermes 3 Llama 3.2 3B Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 1.8 GB | 5.3 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 128K context · ~362 tok/s | |
| 1.9 GB | 5.4 GB | 3-bit small — 3-bit small quantization | Fits: 128K context · ~354 tok/s | |
| 2 GB | 5.5 GB | 3-bit medium — 3-bit medium quantization | Fits: 128K context · ~328 tok/s | |
| 2.0 GB | 5.6 GB | 4-bit · legacy — 4-bit legacy quantization | Fits: 128K context · ~323 tok/s | |
| 2.3 GB | 5.9 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 128K context · ~281 tok/s | |
| 2.7 GB | 6.2 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 128K context · ~245 tok/s | |
| 3.0 GB | 6.5 GB | 6-bit, very good — 6-bit quantization, very good quality | Fits: 128K context · ~218 tok/s | |
| 3.5 GB | 7.1 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 128K context · ~186 tok/s |
Selected Q4_K_M
4-bit medium quantization4.0GBat 16K context
- Weights
- 1.8 GB
- KV cache
- 1.9 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 20.0 GB free · ~281 tok/s
≥50 tok/s needs ≥179 GB/s · RTX 5090 ~500, RTX 3090 Ti ~281
Which Hardware Can Run Hermes 3 Llama 3.2 3B?
Q4_K_M · 2.3 GBHermes 3 Llama 3.2 3B (Q4_K_M) requires 2.3 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 4+ GB is recommended. Using the full 131K context window can add up to 14.8 GB, bringing total usage to 17.1 GB. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 163 devices with unified memory can run Hermes 3 Llama 3.2 3B, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB).
≤ 11 GB
≤ 11 GB tier →fits · 4.0 GB free · up to 32K ctx
16 GPUs · 1 Mac class · 2 device families
RTX 3080~212Simulate output on RTX 3080RTX 2080 Ti~172Simulate output on RTX 2080 TiRTX 3070 Ti~170Simulate output on RTX 3070 TiM3*8 GB~31Simulate output on iMac 24" M3 (8 GB)Orin Nano~29Simulate output on NVIDIA Jetson Orin Nano 8GB (Super)iPhone~21Simulate output on iPhone 1720 GB
fits · 16.0 GB free · up to 128K ctx
1 GPU
RX 7900 XT~206Simulate output on RX 7900 XT32–40 GB
fits · 28.0 GB free · up to 128K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~500Simulate output on RTX 5090M5 Max36 GB~138Simulate output on Mac Studio M5 Max (36 GB)M532 GB~46Simulate output on MacBook Air 13" M5 (32 GB)Orin~57Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~38Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Alternatives to Hermes 3 Llama 3.2 3B
Llama 3 family · 65 models →Similar models
6| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| D1 3BLiquid AI | — | 2.6 GB | ~285 | same size |
| SmolLM3 3BHugging Face | — | 3.4 GB | ~285 | same size |
| Phi 2Microsoft | — | 2.6 GB | ~248 | same size |
| AI21 Jamba Reasoning 3BAI21 Labs | — | 2.3 GB | ~291 | same size |
| Qwen2.5 3B InstructAlibaba | — | 2.8 GB | ~294 | same size |
| Quasar 3B A1B Previewsilx-ai | — | 8.4 GB | ~167 | same size |
Specifications
- Publisher
- Nous Research
- Family
- Llama 3
- Parameters
- 3B
- Architecture
- LlamaForCausalLM
- Context Length
- 131,072 tokens
- Vocabulary Size
- 128,256
- Release Date
- 2024-12-03
- License
- Llama 3 Community
Frequently Asked Questions
How much VRAM does Hermes 3 Llama 3.2 3B need?
Hermes 3 Llama 3.2 3B requires 2.3 GB of VRAM at Q4_K_M, or 6.5 GB at BF16. Full 131K context adds up to 14.8 GB (17.1 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 3B × 4.8 bits ÷ 8 = 1.8 GB
KV Cache + Overhead ≈ 0.5 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 ≈ 15.3 GB (at full 131K context)
VRAM usage by quantization
What's the best quantization for Hermes 3 Llama 3.2 3B?
For Hermes 3 Llama 3.2 3B, Q4_K_M (2.3 GB) offers the best balance of quality and VRAM usage. Q4_K_L (2.4 GB) provides better quality if you have the VRAM. The smallest option is IQ2_M at 1.6 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run Hermes 3 Llama 3.2 3B on a Mac?
Yes — MacBook Air 13" M3 (8 GB) and 145 other Macs can run Hermes 3 Llama 3.2 3B. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 2.3 GB of usable unified memory (RAM minus macOS overhead).
Can I run Hermes 3 Llama 3.2 3B locally?
Yes — Hermes 3 Llama 3.2 3B can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Hermes 3 Llama 3.2 3B?
At Q4_K_M, Hermes 3 Llama 3.2 3B can reach ~2060 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~281 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 ÷ 2.33 × 0.65 = ~2232 tok/s
Estimated speed at Q4_K_M (2.3 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of Hermes 3 Llama 3.2 3B?
At Q4_K_M, the download is about 1.80 GB. The full-precision BF16 version is 6.00 GB. The smallest option (IQ2_M) is 1.01 GB.
Which GPUs can run Hermes 3 Llama 3.2 3B?
52 consumer GPUs can run Hermes 3 Llama 3.2 3B at Q4_K_M. Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT. 52 GPUs have plenty of headroom for comfortable inference.
Which devices can run Hermes 3 Llama 3.2 3B?
163 devices with unified memory can run Hermes 3 Llama 3.2 3B at Q4_K_M (2.3 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.