# OpenAI gpt-oss-120b As of October 22, 2026, OpenAI gpt-oss-120b ranks #72 of 123 models overall with 50.0 ±18.6 on the LLMBenchmark Index. gpt-oss-120b ranks #30 of 108 in math, its strongest category. OpenAI gpt-oss-120b costs $0.10 per million input tokens and $0.50 per million output tokens. gpt-oss-120b was released on August 5, 2025. Page: https://llmbenchmark.ai/models/gpt-oss-120b Data from October 22, 2026. Every figure below names the board it was read from and the day. - Maker: OpenAI - Released: August 5, 2025 - Price per million input tokens: $0.10 - Price per million output tokens: $0.50 - Context window: 131K tokens - Overall rank: 72 of 123, by the LLMBenchmark Index - LLMBenchmark Index: 50.0 ±18.6 (90% margin, from 4 of 7 inputs) - Artificial Analysis Intelligence Index: 11.6 - Vibe Coding Index: not enough data (one index benchmark; at least two are needed to rank) Benchmark results (81 of 195): - Aider polyglot: 41.8% (source: Aider polyglot leaderboard, read October 22, 2026) - SWE-bench Verified: 26.0% (source: SWE-bench (swebench.com), read October 22, 2026) - GPQA Diamond: 78.5% (source: GPQA (Vals AI), read October 22, 2026) - AIME: 92.6% (source: AIME (Vals AI), read October 22, 2026) - MMLU Pro: 79.2% (source: MMLU Pro (Vals AI), read October 22, 2026) - AA Intelligence Index: 11.6 (source: Artificial Analysis (via OpenRouter), read October 22, 2026) - AA Coding Index: 30.4 (source: Artificial Analysis (via OpenRouter), read October 22, 2026) - AA Agentic Index: 3.7 (source: Artificial Analysis (via OpenRouter), read October 22, 2026) - Design Arena: websites: 967 Elo (source: Design Arena, read October 22, 2026) - Design Arena: UI components: 924 Elo (source: Design Arena, read October 22, 2026) - Design Arena: game dev: 992 Elo (source: Design Arena, read October 22, 2026) - Design Arena: data viz: 987 Elo (source: Design Arena, read October 22, 2026) - Design Arena: 3D: 901 Elo (source: Design Arena, read October 22, 2026) - Design Arena: all code categories: 962 Elo (source: Design Arena, read October 22, 2026) - LiveCodeBench (Vals AI): 83.2% (source: Vals AI boards, read October 22, 2026) - SWE-bench Verified (Vals AI): 33.6% (source: Vals AI boards, read October 22, 2026) - MATH 500 (Vals AI): 94.8% (source: Vals AI boards, read October 22, 2026) - MGSM (Vals AI): 92.0% (source: Vals AI boards, read October 22, 2026) - LegalBench (Vals AI): 75.9% (source: Vals AI boards, read October 22, 2026) - CaseLaw (Vals AI): 48.8% (source: Vals AI boards, read October 22, 2026) - CorpFin (Vals AI): 58.2% (source: Vals AI boards, read October 22, 2026) - TaxEval (Vals AI): 71.6% (source: Vals AI boards, read October 22, 2026) - MedQA (Vals AI): 91.4% (source: Vals AI boards, read October 22, 2026) - AA Math Index: 93.4 (source: Artificial Analysis, read October 22, 2026) - GPQA Diamond (Artificial Analysis run): 78.2% (source: Artificial Analysis, read October 22, 2026) - Humanity's Last Exam (Artificial Analysis run): 19.6% (source: Artificial Analysis, read October 22, 2026) - MMLU Pro (Artificial Analysis run): 80.8% (source: Artificial Analysis, read October 22, 2026) - LiveCodeBench (Artificial Analysis run): 87.8% (source: Artificial Analysis, read October 22, 2026) - SciCode: 34.0% (source: Artificial Analysis, read October 22, 2026) - Terminal-Bench Hard (Artificial Analysis run): 23.5% (source: Artificial Analysis, read October 22, 2026) - IFBench: 69.0% (source: Artificial Analysis, read October 22, 2026) - Long context reasoning (AA-LCR): 52.0% (source: Artificial Analysis, read October 22, 2026) - τ²-Bench Telecom: 65.8% (source: Artificial Analysis, read October 22, 2026) - AIME 2025 (Artificial Analysis run): 93.4% (source: Artificial Analysis, read October 22, 2026) - Terminal-Bench: 18.7% (source: Epoch AI benchmarking hub, read October 22, 2026) - WeirdML: 48.2% (source: Epoch AI benchmarking hub, read October 22, 2026) - APEX-Agents: 4.4% (source: Epoch AI benchmarking hub, read October 22, 2026) - METR time horizons: 56.6% (source: Epoch AI benchmarking hub, read October 22, 2026) - GPQA Diamond (Epoch AI run): 75.8% (source: Epoch AI benchmarking hub, read October 22, 2026) - SimpleBench: 22.1% (source: Epoch AI benchmarking hub, read October 22, 2026) - Mystery game puzzles: 2.0% (source: Epoch AI benchmarking hub, read October 22, 2026) - OTIS Mock AIME: 88.9% (source: Epoch AI benchmarking hub, read October 22, 2026) - DTBench: 76.3% (source: Epoch AI benchmarking hub, read October 22, 2026) - LMCA: 22.1% (source: Epoch AI benchmarking hub, read October 22, 2026) - Fiction.LiveBench: 44.4% (source: Epoch AI benchmarking hub, read October 22, 2026) - Creative writing (Lech Mazur): 77.1% (source: Epoch AI benchmarking hub, read October 22, 2026) - Surface Evolver Bench: 25.0% (source: Epoch AI benchmarking hub, read October 22, 2026) - ALE-Bench: 575.6 (source: ALE-Bench (Sakana AI), read October 22, 2026) - Vending-Bench 2: -$22 (source: Epoch AI benchmarking hub, read October 22, 2026) - Epoch Capabilities Index: 139.8 (source: Epoch AI benchmarking hub, read October 22, 2026) - LMArena Text: 1366 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: coding: 1379 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: hard prompts: 1364 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: math: 1389 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: instruction following: 1318 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: creative writing: 1274 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: English: 1375 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: non-English: 1351 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: Chinese: 1385 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: German: 1353 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: Japanese: 1331 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: Korean: 1282 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: French: 1370 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: Spanish: 1389 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: Russian: 1343 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: Polish: 1353 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: multi-turn: 1340 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: longer queries: 1319 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: expert: 1356 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: hard prompts in English: 1368 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: software and IT: 1387 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: mathematical work: 1386 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: science: 1382 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: medicine and healthcare: 1388 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: legal and government: 1355 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: business and finance: 1359 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: writing and language: 1306 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: entertainment and media: 1292 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - Kagi LLM Benchmark: 58.6% (source: Kagi LLM Benchmarking Project, read October 22, 2026) - BrowseComp-Plus: 67.0% (source: BrowseComp-Plus, read October 22, 2026) - SciCode: 7.7% (source: Kaggle Benchmarks, read October 22, 2026)