Vaultgemma 1B — Hardware Requirements & GPU Compatibility
Chat
Vaultgemma 1B is a 1.0B-parameter open language model from Google in the Gemma family. At BF16 it needs about 2.29 GB of VRAM — see which GPUs and Macs can run it below.
4.2K downloads
240 likesSpecifications
- Publisher
- Family
- Gemma
- Parameters
- 1.0B
- Release Date
- 2025-09-12
- License
- Gemma Terms
Get Started
HuggingFace
How Much VRAM Does Vaultgemma 1B Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| BF16 | 16.00 | 2.3 GB | — | 2.08 GB | Brain floating point 16 — preferred for training |
Which GPUs Can Run Vaultgemma 1B?
BF16 · 2.3 GBVaultgemma 1B (BF16) requires 2.3 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 3+ GB is recommended. 35 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti.
Runs great
— Plenty of headroomWhich Devices Can Run Vaultgemma 1B?
BF16 · 2.3 GB33 devices with unified memory can run Vaultgemma 1B, including NVIDIA DGX H100, NVIDIA DGX A100 640GB.
Runs great
— Plenty of headroomRelated Models
Frequently Asked Questions
- How much VRAM does Vaultgemma 1B need?
Vaultgemma 1B requires 2.3 GB of VRAM at BF16.
VRAM = Weights + KV Cache + Overhead
Weights = 1.0B × 16 bits ÷ 8 = 2.1 GB
KV Cache + Overhead ≈ 0.2 GB (at 2K context + ~0.3 GB framework)
VRAM usage by quantization
BF162.3 GB- Can I run Vaultgemma 1B on a Mac?
Vaultgemma 1B requires at least 2.3 GB at BF16, 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 Vaultgemma 1B locally?
Yes — Vaultgemma 1B can run locally on consumer hardware. At BF16 quantization it needs 2.3 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Vaultgemma 1B?
At BF16, Vaultgemma 1B can reach ~1273 tok/s on AMD Instinct MI300X. On NVIDIA GeForce RTX 4090: ~286 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: AMD Instinct MI300X → 5300 ÷ 2.3 × 0.55 = ~1273 tok/s
Estimated speed at BF16 (2.3 GB)
~1273 tok/s~286 tok/s~951 tok/s~787 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Vaultgemma 1B?
At BF16, the download is about 2.08 GB.
- Which GPUs can run Vaultgemma 1B?
35 consumer GPUs can run Vaultgemma 1B at BF16 (2.3 GB). Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT. 35 GPUs have plenty of headroom for comfortable inference.
- Which devices can run Vaultgemma 1B?
33 devices with unified memory can run Vaultgemma 1B at BF16 (2.3 GB), including Mac Mini M4 (16 GB), Mac Mini M4 (32 GB), Mac Mini M4 Pro (24 GB), Mac Mini M4 Pro (48 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.