GPT-OSS 20B
Active Parameters
21B
Context Length
128K
Modality
Text
Architecture
Mixture of Experts
KV Cache Rate
48 KB/tok
License
Apache 2.0
Release Date
5 Aug 2025
Knowledge Cutoff
Jun 2024
API Pricing (per 1M)
Input: $0.02 · Output: $0.11
Hardware Sizing
Precision
Context
Workload Memory: 14.33 GB
INT4 · 8K context
Base Weights: 11.46 GB
KV Cache: 0.38 GB
Runtime & Graphs: 2.50 GB
Minimum Required Hardware
1x NVIDIA L4 24GB
14.3 / 21.6 GB usable
1x RTX 4060 Ti 16GB
14.3 / 14.4 GB usable
Apple M4 24GB
14.3 / 18.0 GB usable
Determine requirements for your specific workload:
Open in VRAM Calculator
Architecture Diagram
Evaluation Benchmarks
Rank
#143
| Benchmark | Score | Rank |
|---|---|---|
Graduate-Level QA | high 0.742 Standard 0.715 | 79 82 |
Agentic Index | high 0.01 | 111 |
Coding Index | high 0.21 | 132 |
General Text | 1317 | 150 |
Intelligence Index | high 0.09 low 0.10 | 248 238 |
Summarization Archived | 0.863 | 6 |
General Knowledge Reference | 0.853 | 11 |
Rankings
Overall Rank
#143
Coding Rank
#138
About GPT-OSS 20B
GPT-OSS 20B is an open-weight Mixture-of-Experts language model by OpenAI designed for efficient local reasoning on consumer hardware. It activates 3.6B parameters per token and features native tool use across a 128K context window.
Technical Specifications
Attention
Attention Structure
Multi-Head Attention
Attention Heads
64
Key-Value Heads
8
Attention Head Dimension
64
Position Embedding
Absolute Position Embedding
RoPE Theta
150,000
Sliding Window Attention
Yes
Sliding Window Size
128
Normalization
RMS Normalization
Activation Function
SwigLU
Dimensions
Hidden Dimension Size
2,880
Number of Layers
24
FFN Intermediate Size (Dense)
2,880
Tokenizer
Vocabulary Size
201,088
Mixture of Experts
Total Expert Parameters
3.6B
Number of Experts
32
Active Experts
4