Gemma 4 31B IT — Hardware Requirements & GPU Compatibility
Vision
Gemma 4 31B IT is a 32.7B-parameter open language model from Google in the Gemma 4 family. It supports a context window of up to 262,144 tokens. At Q4_K_M it needs about 21.23 GB of VRAM — see which GPUs and Macs can run it below.
12.0M downloads
3.3K likesSpecifications
- Publisher
- Family
- Gemma 4
- Parameters
- 32.7B
- Architecture
- Gemma4ForConditionalGeneration
- Context Length
- 262,144 tokens
- Vocabulary Size
- 262,144
- Release Date
- 2026-03-11
- License
- Apache 2.0
Get Started
HuggingFace
How Much VRAM Does Gemma 4 31B IT Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| Q2_K | 3.40 | 15.5 GB | 183.3 GB | 13.89 GB | 2-bit quantization with K-quant improvements |
| Q3_K_S | 3.50 | 15.9 GB | 183.7 GB | 14.30 GB | 3-bit small quantization |
| Q3_K_M | 3.90 | 17.6 GB | 185.3 GB | 15.93 GB | 3-bit medium quantization |
| Q4_0 | 4.00 | 18.0 GB | 185.8 GB | 16.34 GB | 4-bit legacy quantization |
| Q4_K_M | 4.80 | 21.2 GB | 189.0 GB | 19.61 GB | 4-bit medium quantization — most popular sweet spot |
| Q5_K_M | 5.70 | 24.9 GB | 192.7 GB | 23.29 GB | 5-bit medium quantization — good quality/size tradeoff |
| Q6_K | 6.60 | 28.6 GB | 196.4 GB | 26.96 GB | 6-bit quantization, very good quality |
| Q8_0 | 8.00 | 34.3 GB | 202.1 GB | 32.68 GB | 8-bit quantization, near-lossless |
Which GPUs Can Run Gemma 4 31B IT?
Q4_K_M · 21.2 GBGemma 4 31B IT (Q4_K_M) requires 21.2 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 28+ GB is recommended. Using the full 262K context window can add up to 167.8 GB, bringing total usage to 189.0 GB. 7 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti.
Runs great
— Plenty of headroom
Which Devices Can Run Gemma 4 31B IT?
Q4_K_M · 21.2 GB41 devices with unified memory can run Gemma 4 31B IT, including NVIDIA DGX H100, NVIDIA DGX A100 640GB, Mac Mini M4 Pro (24 GB).
Runs great
— Plenty of headroomDecent
— Enough memory, may be tightWhere to Download Gemma 4 31B IT
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 Gemma 4 31B IT need?
Gemma 4 31B IT requires 21.2 GB of VRAM at Q4_K_M, or 67.0 GB at BF16. Full 262K context adds up to 167.8 GB (189.0 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 32.7B × 4.8 bits ÷ 8 = 19.6 GB
KV Cache + Overhead ≈ 1.6 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 169.4 GB (at full 262K context)
VRAM usage by quantization
Q4_K_M21.2 GBQ4_K_M + full context189.0 GB- Can NVIDIA GeForce RTX 4090 run Gemma 4 31B IT?
Yes, at Q4_K_M (21.2 GB) or lower. Higher quantizations like Q5_K_S (24.1 GB) exceed the NVIDIA GeForce RTX 4090's 24 GB.
- What's the best quantization for Gemma 4 31B IT?
For Gemma 4 31B IT, Q4_K_M (21.2 GB) offers the best balance of quality and VRAM usage. Q5_K_S (24.1 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XXS at 10.6 GB.
VRAM requirement by quantization
IQ2_XXS10.6 GBQ3_K_S15.9 GBQ4_120.0 GBQ4_K_M ★21.2 GBQ5_K_S24.1 GBBF1667.0 GB★ Recommended — best balance of quality and VRAM usage.
- Can I run Gemma 4 31B IT on a Mac?
Gemma 4 31B IT requires at least 10.6 GB at IQ2_XXS, 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 Gemma 4 31B IT locally?
Yes — Gemma 4 31B IT can run locally on consumer hardware. At Q4_K_M quantization it needs 21.2 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Gemma 4 31B IT?
At Q4_K_M, Gemma 4 31B IT can reach ~207 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~31 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 ÷ 21.2 × 0.65 = ~245 tok/s
Estimated speed at Q4_K_M (21.2 GB)
~245 tok/s~31 tok/s~245 tok/s~207 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Gemma 4 31B IT?
At Q4_K_M, the download is about 19.61 GB. The full-precision BF16 version is 65.36 GB. The smallest option (IQ2_XXS) is 8.99 GB.
- Which GPUs can run Gemma 4 31B IT?
7 consumer GPUs can run Gemma 4 31B IT at Q4_K_M (21.2 GB). Top options include NVIDIA GeForce RTX 5090, AMD Radeon RX 7900 XTX, NVIDIA GeForce RTX 3090. 1 GPU have plenty of headroom for comfortable inference.
- Which devices can run Gemma 4 31B IT?
41 devices with unified memory can run Gemma 4 31B IT at Q4_K_M (21.2 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Asus ROG Flow Z13 (2025, Ryzen AI Max+ 395, 128 GB), Beelink GTR9 Pro (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.