GPT-OSS
How much VRAM does gpt-oss-20b need?
gpt-oss-20b has 21B parameters, so its weights alone take about 42 GB of VRAM at FP16 — or roughly 11 GB quantized to INT4. Real serving adds KV cache on top, which scales with your context length and concurrent requests. Size the exact figure for your workload below.
OpenAI's smaller open-weight MoE release, natively quantized to MXFP4 for the MoE weights.
Total params
21B
Active params
3.6B
Layers
24
Hidden size
2880
Attention heads
64
KV heads (GQA)
8
Vocab size
201,088
Native context window
131,072
Native precision
BF16
Experts (total)
32
Experts active / token
4
Size gpt-oss-20b for your workload
Opens the free calculator with gpt-oss-20b loaded at BF16 weights and FP16 KV cache. Set your context length and traffic to get exact GPU memory and a ranked list of cloud instances that fit — no signup.
Size this model →