# DeepSeek V3.2 As of October 21, 2026, DeepSeek V3.2 ranks #93 of 213 models overall with 21.5 on the Artificial Analysis Intelligence Index. Its strongest category is knowledge, where it ranks #59 of 147. DeepSeek V3.2 costs $0.28 per million input tokens and $0.42 per million output tokens. It was released on December 1, 2025. Page: https://llmbenchmark.ai/models/deepseek-v3-2 Data from October 21, 2026. Every figure below names the board it was read from and the day. - Maker: DeepSeek - Released: December 1, 2025 - Price per million input tokens: $0.28 - Price per million output tokens: $0.42 - Context window: 164K tokens - Overall rank: 93 of 213, by the Artificial Analysis Intelligence Index - Artificial Analysis Intelligence Index: 21.5 - Vibe Coding Index: 54.4 (from 2 of 4 coding benchmarks) Benchmark results (75 of 190): - SWE-bench Verified: 70.0% (source: SWE-bench (swebench.com), read October 21, 2026) - SWE-bench Multilingual: 59.0% (source: SWE-bench (swebench.com), read October 21, 2026) - LMArena WebDev: 1362 Elo (source: LMArena WebDev, read October 21, 2026) - GPQA Diamond: 80.3% (source: GPQA (Vals AI), read October 21, 2026) - AIME: 84.6% (source: AIME (Vals AI), read October 21, 2026) - MMLU Pro: 84.9% (source: MMLU Pro (Vals AI), read October 21, 2026) - AA Intelligence Index: 21.5 (source: Artificial Analysis (via OpenRouter), read October 21, 2026) - AA Coding Index: 44.2 (source: Artificial Analysis (via OpenRouter), read October 21, 2026) - Design Arena: websites: 1175 Elo (source: Design Arena, read October 21, 2026) - Design Arena: UI components: 1149 Elo (source: Design Arena, read October 21, 2026) - Design Arena: game dev: 1132 Elo (source: Design Arena, read October 21, 2026) - Design Arena: data viz: 1158 Elo (source: Design Arena, read October 21, 2026) - Design Arena: 3D: 1133 Elo (source: Design Arena, read October 21, 2026) - Design Arena: all code categories: 1164 Elo (source: Design Arena, read October 21, 2026) - Design Arena: SVG: 1032 Elo (source: Design Arena, read October 21, 2026) - Design Arena: ASCII art: 1082 Elo (source: Design Arena, read October 21, 2026) - LiveCodeBench (Vals AI): 80.7% (source: Vals AI boards, read October 21, 2026) - SWE-bench Verified (Vals AI): 67.6% (source: Vals AI boards, read October 21, 2026) - Vibe Code Bench (Vals AI): 5.1% (source: Vals AI boards, read October 21, 2026) - MGSM (Vals AI): 86.0% (source: Vals AI boards, read October 21, 2026) - LegalBench (Vals AI): 76.1% (source: Vals AI boards, read October 21, 2026) - CaseLaw (Vals AI): 55.4% (source: Vals AI boards, read October 21, 2026) - CorpFin (Vals AI): 51.0% (source: Vals AI boards, read October 21, 2026) - TaxEval (Vals AI): 68.2% (source: Vals AI boards, read October 21, 2026) - MedQA (Vals AI): 93.9% (source: Vals AI boards, read October 21, 2026) - AA Math Index: 92.0 (source: Artificial Analysis, read October 21, 2026) - GPQA Diamond (Artificial Analysis run): 84.0% (source: Artificial Analysis, read October 21, 2026) - Humanity's Last Exam (Artificial Analysis run): 24.6% (source: Artificial Analysis, read October 21, 2026) - MMLU Pro (Artificial Analysis run): 86.2% (source: Artificial Analysis, read October 21, 2026) - LiveCodeBench (Artificial Analysis run): 86.2% (source: Artificial Analysis, read October 21, 2026) - Terminal-Bench Hard (Artificial Analysis run): 35.6% (source: Artificial Analysis, read October 21, 2026) - IFBench: 60.7% (source: Artificial Analysis, read October 21, 2026) - Long context reasoning (AA-LCR): 73.3% (source: Artificial Analysis, read October 21, 2026) - τ²-Bench Telecom: 90.6% (source: Artificial Analysis, read October 21, 2026) - AIME 2025 (Artificial Analysis run): 92.0% (source: Artificial Analysis, read October 21, 2026) - Terminal-Bench: 39.6% (source: Epoch AI benchmarking hub, read October 21, 2026) - APEX-Agents: 21.3% (source: Epoch AI benchmarking hub, read October 21, 2026) - ARC-AGI: 57.0% (source: Epoch AI benchmarking hub, read October 21, 2026) - ARC-AGI-2: 4.0% (source: Epoch AI benchmarking hub, read October 21, 2026) - DTBench: 85.6% (source: Epoch AI benchmarking hub, read October 21, 2026) - LMCA: 28.8% (source: Epoch AI benchmarking hub, read October 21, 2026) - CL-bench: 12.4% (source: Epoch AI benchmarking hub, read October 21, 2026) - CL-bench Life: 7.4% (source: Epoch AI benchmarking hub, read October 21, 2026) - Vending-Bench 2: $1,034 (source: Epoch AI benchmarking hub, read October 21, 2026) - Epoch Capabilities Index: 146.1 (source: Epoch AI benchmarking hub, read October 21, 2026) - LMArena Text: 1425 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: coding: 1453 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: hard prompts: 1434 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: math: 1435 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: instruction following: 1413 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: creative writing: 1400 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: English: 1439 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: non-English: 1409 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: Chinese: 1455 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: German: 1413 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: Japanese: 1374 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: Korean: 1370 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: French: 1429 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: Spanish: 1424 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: Russian: 1424 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: Polish: 1419 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: multi-turn: 1427 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: longer queries: 1428 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: expert: 1436 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: hard prompts in English: 1446 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: software and IT: 1449 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: mathematical work: 1440 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: science: 1446 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: medicine and healthcare: 1440 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: legal and government: 1425 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: business and finance: 1410 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: writing and language: 1407 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena Text: entertainment and media: 1395 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena WebDev: React: 1350 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026) - LMArena WebDev: HTML: 1371 Elo (source: LMArena (Hugging Face dataset), read October 21, 2026)