# OpenAI GPT-5.4 Mini As of October 22, 2026, OpenAI GPT-5.4 Mini ranks #85 of 123 models overall with 44.0 ±11.5 on the LLMBenchmark Index. GPT-5.4 Mini ranks #57 of 108 in math, its strongest category. OpenAI GPT-5.4 Mini costs $0.75 per million input tokens and $4.50 per million output tokens. GPT-5.4 Mini was released on March 17, 2026. Page: https://llmbenchmark.ai/models/gpt-5-4-mini Data from October 22, 2026. Every figure below names the board it was read from and the day. - Maker: OpenAI - Released: March 17, 2026 - Price per million input tokens: $0.75 - Price per million output tokens: $4.50 - Context window: 400K tokens - Overall rank: 85 of 123, by the LLMBenchmark Index - LLMBenchmark Index: 44.0 ±11.5 (90% margin, from 7 of 7 inputs) - Artificial Analysis Intelligence Index: 24.1 - Vibe Coding Index: 35.8 (from 5 of 6 coding benchmarks) Benchmark results (89 of 195): - LMArena WebDev: 1397 Elo (source: LMArena WebDev, read October 22, 2026) - GPQA Diamond: 83.1% (source: GPQA (Vals AI), read October 22, 2026) - AIME: 95.6% (source: AIME (Vals AI), read October 22, 2026) - MMLU Pro: 84.5% (source: MMLU Pro (Vals AI), read October 22, 2026) - AA Intelligence Index: 24.1 (source: Artificial Analysis (via OpenRouter), read October 22, 2026) - AA Coding Index: 56.1 (source: Artificial Analysis (via OpenRouter), read October 22, 2026) - AA Agentic Index: 17.9 (source: Artificial Analysis (via OpenRouter), read October 22, 2026) - LiveCodeBench (Vals AI): 81.5% (source: Vals AI boards, read October 22, 2026) - SWE-bench Verified (Vals AI): 73.0% (source: Vals AI boards, read October 22, 2026) - Terminal-Bench 4 (Vals AI): 2.5% (source: Vals AI boards, read October 22, 2026) - Vibe Code Bench (Vals AI): 48.0% (source: Vals AI boards, read October 22, 2026) - Code migration (Vals AI): 12.9% (source: Vals AI boards, read October 22, 2026) - ProgramBench (Vals AI): 0.0% (source: Vals AI boards, read October 22, 2026) - Terminal-Bench 2.1 (Vals AI): 54.7% (source: Vals AI boards, read October 22, 2026) - Finance Agent (Vals AI): 45.4% (source: Vals AI boards, read October 22, 2026) - Tax Agent Bench (Vals AI): 8.8% (source: Vals AI boards, read October 22, 2026) - Legal Research Bench (Vals AI): 12.5% (source: Vals AI boards, read October 22, 2026) - Excel Modeling Benchmark (Vals AI): 45.4% (source: Vals AI boards, read October 22, 2026) - MMMU Pro (Vals AI): 79.3% (source: Vals AI boards, read October 22, 2026) - SAGE (Vals AI): 50.8% (source: Vals AI boards, read October 22, 2026) - CaseLaw (Vals AI): 51.7% (source: Vals AI boards, read October 22, 2026) - CorpFin (Vals AI): 60.9% (source: Vals AI boards, read October 22, 2026) - TaxEval (Vals AI): 71.2% (source: Vals AI boards, read October 22, 2026) - MortgageTax (Vals AI): 63.5% (source: Vals AI boards, read October 22, 2026) - Vals Multimodal Index: 54.2% (source: Vals AI boards, read October 22, 2026) - GPQA Diamond (Artificial Analysis run): 87.5% (source: Artificial Analysis, read October 22, 2026) - Humanity's Last Exam (Artificial Analysis run): 28.1% (source: Artificial Analysis, read October 22, 2026) - SciCode: 52.1% (source: Artificial Analysis, read October 22, 2026) - Terminal-Bench Hard (Artificial Analysis run): 52.3% (source: Artificial Analysis, read October 22, 2026) - IFBench: 73.3% (source: Artificial Analysis, read October 22, 2026) - Long context reasoning (AA-LCR): 77.0% (source: Artificial Analysis, read October 22, 2026) - τ²-Bench Telecom: 83.3% (source: Artificial Analysis, read October 22, 2026) - FrontierCode: 27.0% (source: Epoch AI benchmarking hub, read October 22, 2026) - WeirdML: 60.3% (source: Epoch AI benchmarking hub, read October 22, 2026) - DeepResearch Bench: 36.3% (source: Epoch AI benchmarking hub, read October 22, 2026) - GPQA Diamond (Epoch AI run): 86.9% (source: Epoch AI benchmarking hub, read October 22, 2026) - ARC-AGI: 63.7% (source: Epoch AI benchmarking hub, read October 22, 2026) - ARC-AGI-2: 18.9% (source: Epoch AI benchmarking hub, read October 22, 2026) - Chess puzzles: 23.0% (source: Epoch AI benchmarking hub, read October 22, 2026) - Mystery game puzzles: 11.0% (source: Epoch AI benchmarking hub, read October 22, 2026) - OTIS Mock AIME: 88.9% (source: Epoch AI benchmarking hub, read October 22, 2026) - FrontierMath, tiers 1 to 3: 51.2% (source: Epoch AI benchmarking hub, read October 22, 2026) - FrontierMath, tier 4: 9.8% (source: Epoch AI benchmarking hub, read October 22, 2026) - ProofBench: 21.0% (source: Epoch AI benchmarking hub, read October 22, 2026) - SimpleQA Verified: 29.4% (source: Epoch AI benchmarking hub, read October 22, 2026) - DTBench: 80.0% (source: Epoch AI benchmarking hub, read October 22, 2026) - LMCA: 40.8% (source: Epoch AI benchmarking hub, read October 22, 2026) - ALE-Bench: 1188.6 (source: ALE-Bench (Sakana AI), read October 22, 2026) - Epoch Capabilities Index: 148.7 (source: Epoch AI benchmarking hub, read October 22, 2026) - LMArena Text: 1412 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: coding: 1438 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: hard prompts: 1424 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: math: 1418 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: instruction following: 1405 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: creative writing: 1370 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Vision: 1245 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: English: 1412 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: non-English: 1405 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: Chinese: 1446 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: German: 1408 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: Japanese: 1375 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: Korean: 1370 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: French: 1441 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: Spanish: 1405 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: Russian: 1417 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: Polish: 1415 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: multi-turn: 1429 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: longer queries: 1408 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: expert: 1435 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: hard prompts in English: 1420 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: software and IT: 1436 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: mathematical work: 1429 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: science: 1417 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: medicine and healthcare: 1395 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: legal and government: 1414 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: business and finance: 1418 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: writing and language: 1393 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: entertainment and media: 1373 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena WebDev: React: 1386 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena WebDev: HTML: 1419 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Vision: English: 1244 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Vision: Chinese: 1286 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Vision: reading text: 1253 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Vision: diagrams: 1266 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Vision: homework: 1285 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Vision: humor: 1235 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Vision: recognizing things: 1218 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Vision: creative writing: 1222 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - Kagi LLM Benchmark: 37.9% (source: Kagi LLM Benchmarking Project, read October 22, 2026)