GPT OSS Safeguard 20B — Hardware Requirements & GPU Compatibility
Chat
GPT OSS Safeguard 20B is a 21.5B-parameter open language model from OpenAI in the GPT-OSS family. It supports a context window of up to 131,072 tokens. At BF16 it needs about 43.39 GB of VRAM — see which GPUs and Macs can run it below.
43.0K downloads
230 likesSpecifications
- Publisher
- OpenAI
- Family
- GPT-OSS
- Parameters
- 21.5B
- Architecture
- GptOssForCausalLM
- Context Length
- 131,072 tokens
- Vocabulary Size
- 201,088
- Release Date
- 2026-01-14
- License
- Apache 2.0
Get Started
HuggingFace
How Much VRAM Does GPT OSS Safeguard 20B Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| BF16 | 16.00 | 43.4 GB | 47.9 GB | 43.02 GB | Brain floating point 16 — preferred for training |
Which GPUs Can Run GPT OSS Safeguard 20B?
BF16 · 43.4 GBGPT OSS Safeguard 20B (BF16) requires 43.4 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 57+ GB is recommended. Using the full 131K context window can add up to 4.5 GB, bringing total usage to 47.9 GB. No single GPU has enough memory — multi-GPU or cluster setups are needed.
Which Devices Can Run GPT OSS Safeguard 20B?
BF16 · 43.4 GB11 devices with unified memory can run GPT OSS Safeguard 20B, including NVIDIA DGX H100, NVIDIA DGX A100 640GB, Mac Studio M4 Max (64 GB).
Runs great
— Plenty of headroom
Related Models
Frequently Asked Questions
- How much VRAM does GPT OSS Safeguard 20B need?
GPT OSS Safeguard 20B requires 43.4 GB of VRAM at BF16. Full 131K context adds up to 4.5 GB (47.9 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 21.5B × 16 bits ÷ 8 = 43 GB
KV Cache + Overhead ≈ 0.4 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 4.9 GB (at full 131K context)
VRAM usage by quantization
BF1643.4 GBBF16 + full context47.9 GB- Can NVIDIA GeForce RTX 5090 run GPT OSS Safeguard 20B?
No — GPT OSS Safeguard 20B requires at least 43.4 GB at BF16, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
- Can I run GPT OSS Safeguard 20B on a Mac?
GPT OSS Safeguard 20B requires at least 43.4 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 GPT OSS Safeguard 20B locally?
Yes — GPT OSS Safeguard 20B can run locally on consumer hardware. At BF16 quantization it needs 43.4 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is GPT OSS Safeguard 20B?
At BF16, GPT OSS Safeguard 20B can reach ~67 tok/s on AMD Instinct MI300X. 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 ÷ 43.4 × 0.55 = ~67 tok/s
Estimated speed at BF16 (43.4 GB)
~67 tok/s~50 tok/s~42 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of GPT OSS Safeguard 20B?
At BF16, the download is about 43.02 GB.
- Which GPUs can run GPT OSS Safeguard 20B?
No single consumer GPU has enough VRAM to run GPT OSS Safeguard 20B at BF16 (43.4 GB). Multi-GPU or professional hardware is required.
- Which devices can run GPT OSS Safeguard 20B?
11 devices with unified memory can run GPT OSS Safeguard 20B at BF16 (43.4 GB), including Mac Mini M4 Pro (48 GB), Mac Pro M2 Ultra (192 GB), Mac Studio M2 Ultra (192 GB), Mac Studio M4 Max (128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.