gpt-oss-safeguard-20b
gpt-oss-safeguard-120b and gpt-oss-safeguard-20b are safety reasoning models built-upon gpt-oss. With these models, you can classify text content based on safety policies that you provide and perform a suite of foundational safety tasks. These models are intended for safety use cases. For other applications, we recommend using gpt-oss models.
Params
21.5 B
Context
131,072
Downloads 30d
19 K
Likes
285
Download history
tracking started — chart appears after 7 days of snapshots (6 recorded)19 K downloads in the last 30 days
Can you run it?
Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.
| File | Quant | Size | Est. VRAM | Verdict on RTX 4090 · 24 GB |
|---|---|---|---|---|
| model.safetensors | u8 | 13.8 GB | 18.9 GB | ✅ Runs comfortably |
| model.safetensors (bf16, full) | bf16 + 128K ctx | 13.8 GB | 67.3 GB | ❌ Won’t fit |
Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.
Run it
copy-paste, exact tags checked against the Hub$ curl -s https://aimodelscomparison.com/api/v1/models/gpt-oss-safeguard-20b
{
"hf_id": "openai/gpt-oss-safeguard-20b",
"params_b": 21.51,
"context_length": 131072,
"license": { "id": "apache-2.0", "commercial": "yes" },
"downloads_30d": 18787,
"vram_estimates": [
{ "quant": "u8", "gb": 18.9 }
],
"updated_at": "2026-10-19T06:00:24Z"
}
Use via API
via OpenRouter · USD per 1M tokens · synced 2026-10-24$0.075
input / 1M
$0.30
output / 1M
$0.038
cached input / 1M
$0.38
1M in + 1M out
- Context window
- 131,072 tokens
- Max output
- 65,536 tokens
- Providers
- 1
- Cheapest provider
- Groq · $0.075 / $0.30
- Best throughput
- —
- AA Intelligence / Coding / Agentic
- — / — / —
✓ Tool calling
✓ Reasoning
✓ Structured outputs
✗ Image input
✗ File input
✗ Audio input
✓ Prompt caching
Specifications
- Architecture
- GptOssForCausalLM
- Parameters
- 21.5 B
- Tensor type
- U8
- Context length
- 131,072
- Vocabulary
- 201,088
- Layers / heads
- 24 / 64
- Licence
- apache-2.0
- First seen on the Hub
- 2025-09-18
- Base model
- gpt-oss-20b
- Training datasets
- undisclosed
- Added to our catalog
- 2026-10-20
Compare with any text-generation model