OpenAIopen-weight
Released: 2025-08-05Artificial Analysis

gpt-oss-20b

gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for...

Intelligence Index
—/ 100
Artificial Analysis Intelligence Index (independent composite)
Throughput Speed
142tok/sec
Output streaming throughput
Time to First Answer
2.1s
Time to first answer token, including any thinking
Cost to Run the Index
—USD
Artificial Analysis cost of the full Index run

Measured results · Artificial Analysis

Independent evaluations of gpt-oss-20b (Low configuration). Values are reported as published; nothing here is estimated or derived. Snapshot 2026-11-15.

Intelligence Index
—
Cost to run the Index
—
Time per task
—
Omniscience index
—
Reasoning model
Yes
Cached-input price
—

Reasoning & knowledge

Coding & terminal agents

Agentic work

Long context & documents

Reasoning-effort variants evaluated

VariantIntelligence IndexCost to run IndexOutput speed (tok/s)
Low——142.4
High——146.2
Full analysis on Artificial Analysis

Headline Evaluation Breakdown

Six independently measured evaluations (Artificial Analysis)
Math · AIME 202562.3%
Long context · AA-LCR31.0%
Knowledge · Humanity's Last Exam5.3%
Instruction following · IFBench57.8%

Sourced Benchmark Evaluations (9)

Verified performance across authoritative benchmarks with provenance tracking.
BenchmarkStatusScoreTrustDateSource TypeProvenance
AIMENearing Saturation62.3%
75
2026-11-15independentSource ↗
GPQANearing Saturation61.1%
78
2026-11-15independentSource ↗
Humanity's Last ExamActive5.3%
93
2026-11-15independentSource ↗
LiveCodeBenchSaturated65.2%
39
2026-11-15independentSource ↗
IFBenchActive57.8%
93
2026-11-15independentSource ↗
τ²-Bench TelecomActive50.3%
93
2026-11-15independentSource ↗
Terminal-Bench HardActive4.5%
93
2026-11-15independentSource ↗
AA-LCR (Long Context Reasoning)Active31%
93
2026-11-15independentSource ↗
CritPtActive0%
93
2026-11-15independentSource ↗
Local Hardware Execution (Ollama)Hardware Compatible
# 1. Pull and execute model locally with Ollama
ollama run openai-gpt-oss-20b

# 2. Or invoke via local OpenAI-compatible endpoint
curl http://localhost:11434/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai-gpt-oss-20b",
    "messages": [{"role": "user", "content": "Analyze reasoning chains on AIME 2026"}]
  }'

Evaluation & Sourcing Notes

Scores listed for gpt-oss-20b represent verified evaluations extracted from official research papers, independent evaluation suites (HELM, LMSYS, OpenCompass, LiveBench), and verified audit reports.

All benchmarks marked as saturated or deprecated reflect historical performance where the benchmark no longer provides active discriminative power.