# DeepSeek V4 Pro 0813 As of October 22, 2026, DeepSeek V4 Pro 0813 ranks #38 of 123 models overall with 60.8 ±8.5 on the LLMBenchmark Index. DeepSeek V4 Pro 0813 ranks #37 of 108 in math, its strongest category. DeepSeek V4 Pro 0813 costs $0.66 per million input tokens and $1.98 per million output tokens. DeepSeek V4 Pro 0813 was released on August 12, 2026. Page: https://llmbenchmark.ai/models/deepseek-v4-pro-0813 Data from October 22, 2026. Every figure below names the board it was read from and the day. - Maker: DeepSeek - Released: August 12, 2026 - Price per million input tokens: $0.66 - Price per million output tokens: $1.98 - Context window: 1.05M tokens - Overall rank: 38 of 123, by the LLMBenchmark Index - LLMBenchmark Index: 60.8 ±8.5 (90% margin, from 7 of 7 inputs) - Artificial Analysis Intelligence Index: 36.0 - Vibe Coding Index: 56.1 (from 5 of 6 coding benchmarks) Benchmark results (78 of 195): - LMArena WebDev: 1582 Elo (source: LMArena WebDev, read October 22, 2026) - Vals Index: 53.0% (source: Vals Index, read October 22, 2026) - GPQA Diamond: 92.4% (source: GPQA (Vals AI), read October 22, 2026) - MMLU Pro: 87.0% (source: MMLU Pro (Vals AI), read October 22, 2026) - AA Intelligence Index: 36.0 (source: Artificial Analysis (via OpenRouter), read October 22, 2026) - AA Coding Index: 68.8 (source: Artificial Analysis (via OpenRouter), read October 22, 2026) - AA Agentic Index: 41.3 (source: Artificial Analysis (via OpenRouter), read October 22, 2026) - LiveCodeBench (Vals AI): 87.5% (source: Vals AI boards, read October 22, 2026) - SWE-bench Verified (Vals AI): 96.4% (source: Vals AI boards, read October 22, 2026) - Terminal-Bench 4 (Vals AI): 14.1% (source: Vals AI boards, read October 22, 2026) - Vibe Code Bench (Vals AI): 82.3% (source: Vals AI boards, read October 22, 2026) - Code migration (Vals AI): 41.5% (source: Vals AI boards, read October 22, 2026) - ProgramBench (Vals AI): 0.0% (source: Vals AI boards, read October 22, 2026) - SREBench (Vals AI): 1.9% (source: Vals AI boards, read October 22, 2026) - IOI (Vals AI): 51.6% (source: Vals AI boards, read October 22, 2026) - Terminal-Bench 2.1 (Vals AI): 54.7% (source: Vals AI boards, read October 22, 2026) - Vibe Code Bench 1-100 (Vals AI): 17.5% (source: Vals AI boards, read October 22, 2026) - SkillsBench (Vals AI): 53.8% (source: Vals AI boards, read October 22, 2026) - CyberBench (Vals AI): 72.9% (source: Vals AI boards, read October 22, 2026) - Finance Agent (Vals AI): 50.4% (source: Vals AI boards, read October 22, 2026) - Tax Agent Bench (Vals AI): 26.8% (source: Vals AI boards, read October 22, 2026) - Legal Research Bench (Vals AI): 40.9% (source: Vals AI boards, read October 22, 2026) - Harvey's Legal Agent Benchmark (Vals AI): 10.8% (source: Vals AI boards, read October 22, 2026) - Excel Modeling Benchmark (Vals AI): 52.8% (source: Vals AI boards, read October 22, 2026) - Terminal-Bench-Science (Vals AI): 4.3% (source: Vals AI boards, read October 22, 2026) - LegalBench (Vals AI): 82.4% (source: Vals AI boards, read October 22, 2026) - CorpFin (Vals AI): 65.4% (source: Vals AI boards, read October 22, 2026) - TaxEval (Vals AI): 73.1% (source: Vals AI boards, read October 22, 2026) - MedCode (Vals AI): 42.5% (source: Vals AI boards, read October 22, 2026) - MedScribe (Vals AI): 80.2% (source: Vals AI boards, read October 22, 2026) - GPQA Diamond (Artificial Analysis run): 92.8% (source: Artificial Analysis, read October 22, 2026) - Humanity's Last Exam (Artificial Analysis run): 41.0% (source: Artificial Analysis, read October 22, 2026) - SciCode: 51.0% (source: Artificial Analysis, read October 22, 2026) - Long context reasoning (AA-LCR): 80.3% (source: Artificial Analysis, read October 22, 2026) - FrontierCode: 28.5% (source: Epoch AI benchmarking hub, read October 22, 2026) - WeirdML: 66.2% (source: Epoch AI benchmarking hub, read October 22, 2026) - APEX-Agents: 47.3% (source: Epoch AI benchmarking hub, read October 22, 2026) - GPQA Diamond (Epoch AI run): 91.7% (source: Epoch AI benchmarking hub, read October 22, 2026) - ARC-AGI: 90.5% (source: Epoch AI benchmarking hub, read October 22, 2026) - ARC-AGI-2: 61.3% (source: Epoch AI benchmarking hub, read October 22, 2026) - Chess puzzles: 46.0% (source: Epoch AI benchmarking hub, read October 22, 2026) - Mystery game puzzles: 43.0% (source: Epoch AI benchmarking hub, read October 22, 2026) - OTIS Mock AIME: 98.6% (source: Epoch AI benchmarking hub, read October 22, 2026) - FrontierMath, tiers 1 to 3: 64.6% (source: Epoch AI benchmarking hub, read October 22, 2026) - FrontierMath, tier 4: 26.8% (source: Epoch AI benchmarking hub, read October 22, 2026) - ProofBench: 50.0% (source: Epoch AI benchmarking hub, read October 22, 2026) - SimpleQA Verified: 52.9% (source: Epoch AI benchmarking hub, read October 22, 2026) - DTBench: 93.9% (source: Epoch AI benchmarking hub, read October 22, 2026) - LMCA: 45.5% (source: Epoch AI benchmarking hub, read October 22, 2026) - ALE-Bench: 1403.2 (source: ALE-Bench (Sakana AI), read October 22, 2026) - Epoch Capabilities Index: 155.6 (source: Epoch AI benchmarking hub, read October 22, 2026) - LMArena Text: 1446 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: coding: 1469 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: hard prompts: 1458 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: math: 1441 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: instruction following: 1448 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: creative writing: 1439 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: English: 1457 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: non-English: 1431 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: Chinese: 1481 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: Korean: 1449 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: French: 1446 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: Spanish: 1435 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: Russian: 1448 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: multi-turn: 1454 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: longer queries: 1458 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: expert: 1459 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: hard prompts in English: 1463 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: software and IT: 1462 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: mathematical work: 1443 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: science: 1457 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: medicine and healthcare: 1446 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: legal and government: 1459 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: business and finance: 1442 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: writing and language: 1443 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena Text: entertainment and media: 1424 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena WebDev: React: 1588 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026) - LMArena WebDev: HTML: 1566 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)