OpenAIopen-weight
Released: 2025-08-05Artificial Analysisgpt-oss-120b
gpt-oss-120b is an open-weight, 117B-parameter Mixture-of-Experts (MoE) language model from OpenAI designed for high-reasoning, agentic, and general-purpose production use cases. It activates 5.1B parameters per forward pass and is optimized...
Intelligence Index
3.8/ 100
Artificial Analysis Intelligence Index (independent composite)Throughput Speed
167tok/sec
Output streaming throughputTime to First Answer
17s
Time to first answer token, including any thinkingCost to Run the Index
0.35USD
Artificial Analysis cost of the full Index runMeasured results · Artificial Analysis
Independent evaluations of gpt-oss-120b (High configuration). Values are reported as published; nothing here is estimated or derived. Snapshot 2026-11-15.
- Intelligence Index
- 3.8
- Cost to run the Index
- $0.35
- Time per task
- 7.3 min
- Omniscience index
- -46.4
- Reasoning model
- Yes
- Cached-input price
- $0.125 / 1M
Reasoning & knowledge
- AA-Omniscience Accuracy21.7%
- AIME 202593.4%
- CritPt1.1%
- CritPt v2 (Hard)4.0%
- GPQA Diamond78.2%
- HLE-Diamond3.8%
- Humanity's Last Exam19.6%
- IFBench69.0%
Coding & terminal agents
- FrontierCode 1.11.2%
- FrontierCode-AA 2.00.0%
- LiveCodeBench87.8%
- SciCode34.0%
- Terminal-Bench 2.126.2%
- Terminal-Bench 4.00.0%
- Terminal-Bench Hard23.5%
- Terminal-Bench-Science 0.10.0%
Agentic work
- AA-Briefcase v1.1 (Elo)0 Elo
- APEX-Agents-AA3.1%
- AutomationBench-AA0.2%
- EnterpriseOps-Gym-AA25.5%
- GDPval-AA (normalized Elo)615 Elo
- ITBench-AA5.7%
- τ²-Bench Telecom65.8%
- τ³-Banking12.8%
Long context & documents
Health & safety
Reasoning-effort variants evaluated
| Variant | Intelligence Index | Cost to run Index | Output speed (tok/s) |
|---|---|---|---|
| High | 3.8 | $0.35 | 167.0 |
| Low | — | — | 174.5 |
Headline Evaluation Breakdown
Six independently measured evaluations (Artificial Analysis)Reasoning · CritPt v2 (Hard)4.0%
Math · AIME 202593.4%
Long context · AA-LCR52.0%
Knowledge · Humanity's Last Exam19.6%
Instruction following · IFBench69.0%
Sourced Benchmark Evaluations (26)
Verified performance across authoritative benchmarks with provenance tracking.| Benchmark | Status | Score | Trust | Date | Source Type | Provenance |
|---|---|---|---|---|---|---|
| AA-LCR (Long Context Reasoning) | Active | 52% | 93 | 2026-11-15 | independent | Source ↗ |
| APEX-Agents-AA | Active | 3.1% | 93 | 2026-11-15 | independent | Source ↗ |
| ITBench-AA | Active | 5.6% | 93 | 2026-11-15 | independent | Source ↗ |
| EnterpriseOps-Gym-AA | Active | 25.5% | 93 | 2026-11-15 | independent | Source ↗ |
| AutomationBench-AA | Active | 0.2% | 93 | 2026-11-15 | independent | Source ↗ |
| MLCR-AA (Medical Long Context Reasoning) | Active | 1.7% | 93 | 2026-11-15 | independent | Source ↗ |
| GDP.pdf | Active | 4.2% | 93 | 2026-11-15 | independent | Source ↗ |
| FrontierCode-AA 2.0 | Active | 0% | 93 | 2026-11-15 | independent | Source ↗ |
| AA-Omniscience (Non-Hallucination) | Active | 13.1% | 93 | 2026-11-15 | independent | Source ↗ |
| AA-Omniscience (Accuracy) | Active | 21.7% | 93 | 2026-11-15 | independent | Source ↗ |
| HLE-Diamond | Active | 3.8% | 93 | 2026-11-15 | independent | Source ↗ |
| CritPt v2 (Hard) | Active | 4% | 93 | 2026-11-15 | independent | Source ↗ |
| CritPt | Active | 1.1% | 93 | 2026-11-15 | independent | Source ↗ |
| AIME | Nearing Saturation | 93.4% | 75 | 2026-11-15 | independent | Source ↗ |
| Terminal-Bench-Science 0.1 | Active | 0% | 93 | 2026-11-15 | independent | Source ↗ |
| Terminal-Bench 4.0 | Active | 0% | 93 | 2026-11-15 | independent | Source ↗ |
| Terminal-Bench 2.1 | Active | 26.2% | 93 | 2026-11-15 | independent | Source ↗ |
| Terminal-Bench Hard | Active | 23.5% | 93 | 2026-11-15 | independent | Source ↗ |
| τ³-Banking | Active | 12.8% | 93 | 2026-11-15 | independent | Source ↗ |
| τ²-Bench Telecom | Active | 65.8% | 93 | 2026-11-15 | independent | Source ↗ |
| SciCode | Active | 34% | 93 | 2026-11-15 | independent | Source ↗ |
| IFBench | Active | 69% | 93 | 2026-11-15 | independent | Source ↗ |
| LiveCodeBench | Saturated | 87.8% | 39 | 2026-11-15 | independent | Source ↗ |
| Humanity's Last Exam | Active | 19.6% | 93 | 2026-11-15 | independent | Source ↗ |
| GPQA | Nearing Saturation | 78.2% | 78 | 2026-11-15 | independent | Source ↗ |
| FrontierCode | Active | 1.2% | 98 | 2026-11-15 | independent | Source ↗ |
Local Hardware Execution (Ollama)Hardware Compatible
# 1. Pull and execute model locally with Ollama
ollama run openai-gpt-oss-120b
# 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-120b",
"messages": [{"role": "user", "content": "Analyze reasoning chains on AIME 2026"}]
}'Evaluation & Sourcing Notes
Scores listed for gpt-oss-120b 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.