# Anthropic Claude Sonnet 4.5 As of October 21, 2026, Anthropic Claude Sonnet 4.5 ranks #75 of 123 models overall with 49.5 ±9.5 on the LLMBenchmark Index. Claude Sonnet 4.5 ranks #75 of 133 in web design, its strongest category. Anthropic Claude Sonnet 4.5 costs $3.00 per million input tokens and $15.00 per million output tokens. Claude Sonnet 4.5 was released on September 29, 2025. Page: https://llmbenchmark.ai/models/claude-sonnet-4-5 Data from October 21, 2026. Every figure below names the board it was read from and the day. - Maker: Anthropic - Released: September 29, 2025 - Price per million input tokens: $3.00 - Price per million output tokens: $15.00 - Context window: 1M tokens - Overall rank: 75 of 123, by the LLMBenchmark Index - LLMBenchmark Index: 49.5 ±9.5 (90% margin, from 6 of 7 inputs) - Artificial Analysis Intelligence Index: 20.7 - Vibe Coding Index: 39.5 (from 3 of 6 coding benchmarks) Benchmark results (95 of 195): - SWE-bench Verified: 74.8% (source: SWE-bench (swebench.com), read October 21, 2026) - SWE-bench Multilingual: 67.0% (source: SWE-bench (swebench.com), read October 21, 2026) - LMArena WebDev: 1394 Elo (source: LMArena WebDev, read October 21, 2026) - GPQA Diamond: 81.6% (source: GPQA (Vals AI), read October 21, 2026) - AIME: 88.2% (source: AIME (Vals AI), read October 21, 2026) - MMLU Pro: 87.4% (source: MMLU Pro (Vals AI), read October 21, 2026) - AA Intelligence Index: 20.7 (source: Artificial Analysis (via OpenRouter), read October 21, 2026) - AA Coding Index: 52.1 (source: Artificial Analysis (via OpenRouter), read October 21, 2026) - AA Agentic Index: 15.8 (source: Artificial Analysis (via OpenRouter), read October 21, 2026) - Design Arena: websites: 1190 Elo (source: Design Arena, read October 21, 2026) - Design Arena: UI components: 1173 Elo (source: Design Arena, read October 21, 2026) - Design Arena: game dev: 1166 Elo (source: Design Arena, read October 21, 2026) - Design Arena: data viz: 1167 Elo (source: Design Arena, read October 21, 2026) - Design Arena: 3D: 1158 Elo (source: Design Arena, read October 21, 2026) - Design Arena: all code categories: 1183 Elo (source: Design Arena, read October 21, 2026) - Design Arena: SVG: 1115 Elo (source: Design Arena, read October 21, 2026) - Design Arena: ASCII art: 1205 Elo (source: Design Arena, read October 21, 2026) - Design Arena: agents building web apps: 1059 Elo (source: Design Arena, read October 21, 2026) - Design Arena: agents building full-stack apps: 1050 Elo (source: Design Arena, read October 21, 2026) - Design Arena: agents building mobile apps: 1137 Elo (source: Design Arena, read October 21, 2026) - LiveCodeBench (Vals AI): 73.0% (source: Vals AI boards, read October 21, 2026) - SWE-bench Verified (Vals AI): 70.0% (source: Vals AI boards, read October 21, 2026) - Vibe Code Bench (Vals AI): 22.6% (source: Vals AI boards, read October 21, 2026) - MMMU Pro (Vals AI): 79.3% (source: Vals AI boards, read October 21, 2026) - MGSM (Vals AI): 94.3% (source: Vals AI boards, read October 21, 2026) - SAGE (Vals AI): 36.1% (source: Vals AI boards, read October 21, 2026) - LegalBench (Vals AI): 84.1% (source: Vals AI boards, read October 21, 2026) - CaseLaw (Vals AI): 62.2% (source: Vals AI boards, read October 21, 2026) - CorpFin (Vals AI): 62.0% (source: Vals AI boards, read October 21, 2026) - TaxEval (Vals AI): 73.3% (source: Vals AI boards, read October 21, 2026) - MortgageTax (Vals AI): 64.0% (source: Vals AI boards, read October 21, 2026) - MedQA (Vals AI): 94.7% (source: Vals AI boards, read October 21, 2026) - MedCode (Vals AI): 44.1% (source: Vals AI boards, read October 21, 2026) - MedScribe (Vals AI): 84.5% (source: Vals AI boards, read October 21, 2026) - Terminal-Bench: 46.5% (source: Epoch AI benchmarking hub, read October 21, 2026) - GSO-Bench: 14.7% (source: Epoch AI benchmarking hub, read October 21, 2026) - WeirdML: 47.7% (source: Epoch AI benchmarking hub, read October 21, 2026) - METR time horizons: 67.4% (source: Epoch AI benchmarking hub, read October 21, 2026) - Remote Labor Index: 2.1% (source: Epoch AI benchmarking hub, read October 21, 2026) - GDPval: 42.5% (source: Epoch AI benchmarking hub, read October 21, 2026) - DeepResearch Bench: 52.6% (source: Epoch AI benchmarking hub, read October 21, 2026) - GPQA Diamond (Epoch AI run): 82.3% (source: Epoch AI benchmarking hub, read October 21, 2026) - ARC-AGI: 63.7% (source: Epoch AI benchmarking hub, read October 21, 2026) - ARC-AGI-2: 13.6% (source: Epoch AI benchmarking hub, read October 21, 2026) - SimpleBench: 54.3% (source: Epoch AI benchmarking hub, read October 21, 2026) - Chess puzzles: 10.0% (source: Epoch AI benchmarking hub, read October 21, 2026) - Mystery game puzzles: 17.0% (source: Epoch AI benchmarking hub, read October 21, 2026) - EBR-bench: 2.4% (source: Epoch AI benchmarking hub, read October 21, 2026) - OTIS Mock AIME: 77.8% (source: Epoch AI benchmarking hub, read October 21, 2026) - FrontierMath, tiers 1 to 3: 23.9% (source: Epoch AI benchmarking hub, read October 21, 2026) - FrontierMath, tier 4: 2.4% (source: Epoch AI benchmarking hub, read October 21, 2026) - ProofBench: 19.0% (source: Epoch AI benchmarking hub, read October 21, 2026) - MATH level 5: 97.7% (source: Epoch AI benchmarking hub, read October 21, 2026) - SimpleQA Verified: 30.7% (source: Epoch AI benchmarking hub, read October 21, 2026) - DTBench: 83.2% (source: Epoch AI benchmarking hub, read October 21, 2026) - LMCA: 38.8% (source: Epoch AI benchmarking hub, read October 21, 2026) - Cybench: 60.0% (source: Epoch AI benchmarking hub, read October 21, 2026) - OSWorld: 62.9% (source: Epoch AI benchmarking hub, read October 21, 2026) - VPCT: 39.8% (source: Epoch AI benchmarking hub, read October 21, 2026) - Vending-Bench 2: $3,839 (source: Epoch AI benchmarking hub, read October 21, 2026) - Epoch Capabilities Index: 146.7 (source: Epoch AI benchmarking hub, read October 21, 2026) - LMArena Text: 1438 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: coding: 1488 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: hard prompts: 1462 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: math: 1449 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: instruction following: 1459 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: creative writing: 1442 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: English: 1450 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: non-English: 1425 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: Chinese: 1460 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: German: 1427 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: Japanese: 1390 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: Korean: 1402 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: French: 1459 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: Spanish: 1458 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: Russian: 1437 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: Polish: 1423 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: multi-turn: 1465 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: longer queries: 1475 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: expert: 1483 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: hard prompts in English: 1471 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: software and IT: 1473 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: mathematical work: 1462 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: science: 1448 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: medicine and healthcare: 1444 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: legal and government: 1451 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: business and finance: 1440 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: writing and language: 1442 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: entertainment and media: 1428 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena WebDev: React: 1386 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena WebDev: HTML: 1399 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Document: 1450 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - Berkeley Function Calling Leaderboard (BFCL): 73.2% (source: Berkeley Function Calling Leaderboard, read October 21, 2026) - Kagi LLM Benchmark: 57.9% (source: Kagi LLM Benchmarking Project, read October 21, 2026) - FACTS Benchmark Suite: 40.6% (source: Kaggle Benchmarks, read October 21, 2026)