FrankenGemma4 — Hardware Requirements & GPU Compatibility
Chat
FrankenGemma4 is a 1.2B-parameter open language model from stamsam in the Gemma family. It supports a context window of up to 131,072 tokens. At BF16 it needs about 2.87 GB of VRAM — see which GPUs and Macs can run it below.
48 downloads
6 likesSpecifications
- Publisher
- stamsam
- Family
- Gemma
- Parameters
- 1.2B
- Architecture
- Gemma4ForConditionalGeneration
- Context Length
- 131,072 tokens
- Vocabulary Size
- 262,144
- Release Date
- 2026-04-20
- License
- Gemma Terms
Get Started
HuggingFace
How Much VRAM Does FrankenGemma4 Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| BF16 | 16.00 | 2.9 GB | 16.7 GB | 2.35 GB | Brain floating point 16 — preferred for training |
Which GPUs Can Run FrankenGemma4?
BF16 · 2.9 GBFrankenGemma4 (BF16) requires 2.9 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 13.9 GB, bringing total usage to 16.7 GB. 35 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti.
Runs great
— Plenty of headroomWhich Devices Can Run FrankenGemma4?
BF16 · 2.9 GB33 devices with unified memory can run FrankenGemma4, including NVIDIA DGX H100, NVIDIA DGX A100 640GB.
Runs great
— Plenty of headroomRelated Models
Frequently Asked Questions
- How much VRAM does FrankenGemma4 need?
FrankenGemma4 requires 2.9 GB of VRAM at BF16. Full 131K context adds up to 13.9 GB (16.7 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 1.2B × 16 bits ÷ 8 = 2.4 GB
KV Cache + Overhead ≈ 0.5 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 14.3 GB (at full 131K context)
VRAM usage by quantization
BF162.9 GBBF16 + full context16.7 GB- Can I run FrankenGemma4 on a Mac?
FrankenGemma4 requires at least 2.9 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 FrankenGemma4 locally?
Yes — FrankenGemma4 can run locally on consumer hardware. At BF16 quantization it needs 2.9 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is FrankenGemma4?
At BF16, FrankenGemma4 can reach ~1016 tok/s on AMD Instinct MI300X. On NVIDIA GeForce RTX 4090: ~228 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 ÷ 2.9 × 0.55 = ~1016 tok/s
Estimated speed at BF16 (2.9 GB)
~1016 tok/s~228 tok/s~759 tok/s~628 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of FrankenGemma4?
At BF16, the download is about 2.35 GB.
- Which GPUs can run FrankenGemma4?
35 consumer GPUs can run FrankenGemma4 at BF16 (2.9 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 FrankenGemma4?
33 devices with unified memory can run FrankenGemma4 at BF16 (2.9 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.