Rankings / DeepSeek
Deepseek V3 2 Exp
released 2025-09-29
BenchAtlas Index
as of 2026-07-14
thinking
rank #204
9 families · 5 categories · high
no reasoning
rank #242
9 families · 5 categories · high
Benchmark evidence
47 results
agentic coding
| τ²-Bench | no reasoning | 33.9 % | independent | Artificial Analysis |
| τ²-Bench | thinking | 33.9 % | independent | Artificial Analysis |
coding
| LiveCodeBench v6 implementation=artificial-analysis | thinking | 78.9 % | independent | Artificial Analysis |
| LiveCodeBench v6 implementation=artificial-analysis | no reasoning | 55.4 % | independent | Artificial Analysis |
| SciCode | no reasoning | 39.9 % | independent | Artificial Analysis |
| SciCode | thinking | 37.7 % | independent | Artificial Analysis |
external indices
| Intelligence Index v4.1 | thinking | 25.4 points | independent | Artificial Analysis |
| Intelligence Index v4.1 | no reasoning | 21.3 points | independent | Artificial Analysis |
| AA Math Index | thinking | 87.7 points | independent | Artificial Analysis |
| AA Math Index | no reasoning | 57.7 points | independent | Artificial Analysis |
| ECI | 144.9 points | independent | Epoch AI Benchmarking Hub | |
| ECI | high effort | 144.9 points | independent | Epoch AI Benchmarking Hub |
human preference
| Chinese (style control)older version arena=text · category=chinese · style_control=true | 1475.0 1452.3–1497.8 | community | LMArena Leaderboard Dataset | |
| Coding (style control)older version arena=text · category=coding · style_control=true | 1465.3 1453.5–1477.0 | community | LMArena Leaderboard Dataset | |
| Hard Prompts English (style control)older version arena=text · category=hard_prompts_english · style_control=true | 1453.8 1443.3–1464.3 | community | LMArena Leaderboard Dataset | |
| Industry Software And It Services (style control)older version arena=text · category=industry_software_and_it_services · style_control=true | 1453.4 1443.8–1462.9 | community | LMArena Leaderboard Dataset | |
| Hard Prompts (style control)older version arena=text · category=hard_prompts · style_control=true | 1447.5 1439.5–1455.5 | community | LMArena Leaderboard Dataset | |
| Longer Query (style control)older version arena=text · category=longer_query · style_control=true | 1441.9 1431.1–1452.7 | community | LMArena Leaderboard Dataset | |
| English (style control)older version arena=text · category=english · style_control=true | 1437.9 1429.5–1446.3 | community | LMArena Leaderboard Dataset | |
| Industry Medicine And Healthcare (style control)older version arena=text · category=industry_medicine_and_healthcare · style_control=true | 1437.9 1413.9–1461.9 | community | LMArena Leaderboard Dataset | |
| Industry Life And Physical And Social Science (style control)older version arena=text · category=industry_life_and_physical_and_social_science · style_control=true | 1435.2 1421.7–1448.7 | community | LMArena Leaderboard Dataset | |
| Multi Turn (style control)older version arena=text · category=multi_turn · style_control=true | 1430.8 1417.7–1443.8 | community | LMArena Leaderboard Dataset | |
| Overall (style control) arena=text · category=overall · style_control=true | 1422.9 1416.5–1429.3 | community | LMArena Leaderboard Dataset | |
| Expert (style control)older version arena=text · category=expert · style_control=true | 1422.7 1399.3–1446.0 | community | LMArena Leaderboard Dataset | |
| Industry Business And Management And Financial Operations (style control)older version arena=text · category=industry_business_and_management_and_financial_operations · style_control=true | 1422.2 1409.5–1434.8 | community | LMArena Leaderboard Dataset | |
| Math (style control)older version arena=text · category=math · style_control=true | 1417.1 1396.5–1437.7 | community | LMArena Leaderboard Dataset | |
| Industry Legal And Government (style control)older version arena=text · category=industry_legal_and_government · style_control=true | 1416.6 1395.9–1437.3 | community | LMArena Leaderboard Dataset | |
| Instruction Following (style control)older version arena=text · category=instruction_following · style_control=true | 1416.3 1406.0–1426.7 | community | LMArena Leaderboard Dataset | |
| Exclude Ties (style control)older version arena=text · category=exclude_ties · style_control=true | 1416.1 1407.2–1425.0 | community | LMArena Leaderboard Dataset | |
| Industry Writing And Literature And Language (style control)older version arena=text · category=industry_writing_and_literature_and_language · style_control=true | 1413.0 1401.4–1424.6 | community | LMArena Leaderboard Dataset | |
| Industry Mathematical (style control)older version arena=text · category=industry_mathematical · style_control=true | 1412.8 1390.1–1435.5 | community | LMArena Leaderboard Dataset | |
| Creative Writing (style control)older version arena=text · category=creative_writing · style_control=true | 1410.2 1395.5–1424.9 | community | LMArena Leaderboard Dataset | |
| Industry Entertainment And Sports And Media (style control)older version arena=text · category=industry_entertainment_and_sports_and_media · style_control=true | 1407.8 1395.1–1420.4 | community | LMArena Leaderboard Dataset | |
| Non English (style control)older version arena=text · category=non_english · style_control=true | 1406.3 1398.2–1414.5 | community | LMArena Leaderboard Dataset | |
| Russian (style control)older version arena=text · category=russian · style_control=true | 1404.8 1385.1–1424.5 | community | LMArena Leaderboard Dataset |
knowledge science
| GPQA Diamond implementation=artificial-analysis | thinking | 79.7 % | independent | Artificial Analysis |
| GPQA Diamond implementation=artificial-analysis | no reasoning | 73.8 % | independent | Artificial Analysis |
| Humanity's Last Exam implementation=artificial-analysis | thinking | 13.8 % | independent | Artificial Analysis |
| Humanity's Last Exam implementation=artificial-analysis | no reasoning | 8.6 % | independent | Artificial Analysis |
| MMLU-Pro implementation=artificial-analysis | thinking | 85.0 % | independent | Artificial Analysis |
| MMLU-Pro implementation=artificial-analysis | no reasoning | 83.6 % | independent | Artificial Analysis |
long context instruction
| AA-LCR | thinking | 69.0 % | independent | Artificial Analysis |
| AA-LCR | no reasoning | 43.0 % | independent | Artificial Analysis |
| IFBench | thinking | 54.1 % | independent | Artificial Analysis |
| IFBench | no reasoning | 43.1 % | independent | Artificial Analysis |
reasoning math
| AIME year=2025 · implementation=artificial-analysis | thinking | 87.7 % | independent | Artificial Analysis |
| AIME year=2025 · implementation=artificial-analysis | no reasoning | 57.7 % | independent | Artificial Analysis |
Agent + model results
systems, not bare-model scores
| agent + model Artificial Analysis harness + Deepseek V3 2 Exp | Terminal-Bench Hard | 31.1 % | independent | Artificial Analysis |
| agent + model Artificial Analysis harness + Deepseek V3 2 Exp | Terminal-Bench Hard | 25.0 % | independent | Artificial Analysis |
These scores measure the whole agent system (scaffold, tools, budgets) — they are never merged into the bare model’s numbers.