Back to Models

Gpt oss 20b

GPT-OSS-20B is a compact open-weight MoE model from OpenAI containing 21B parameters (3.6B active). It uses grouped multi-query attention for low-latency inference on consumer hardware under an Apache 2.0 license.

Thinking Mode
Parameters
21000000000 B
Context
131,000 tokens
Released
May 8, 2025

Leaderboards

Average Score combining domain-specific Autobench scores; Higher is better

Performance vs. Industry Average

Intelligence

Gpt oss 20b is of lower intelligence compared to average (2.9), with an intelligence score of 2.6.

Price

Gpt oss 20b is cheaper compared to average ($0.75 per 1M Tokens) with a price of $0.01 per 1M Tokens.

Latency

Gpt oss 20b has a lower average latency compared to average (44.25s), with an average latency of 42.87s.

P99 Latency

Gpt oss 20b has a higher P99 latency compared to average (126.46s), taking 165.52s to receive the first token at P99 (TTFT).

Context Window

Gpt oss 20b has a smaller context window than average (406k tokens), with a context window of 131k tokens.

Gpt oss 20b - AutoBench