Rankings / DeepSeek
Deepseek V3 1
released 2025-08-21
BenchAtlas Index
as of 2026-09-11
thinking
rank #258
9 families · 5 categories · high
no reasoning
rank #313
9 families · 5 categories · high
Benchmark evidence
48 results
agentic coding
| τ²-Bench | thinking | 37.4 % | independent | Artificial Analysis |
| τ²-Bench | no reasoning | 34.8 % | independent | Artificial Analysis |
coding
| LiveCodeBench v6 implementation=artificial-analysis | thinking | 78.4 % | independent | Artificial Analysis |
| LiveCodeBench v6 implementation=artificial-analysis | no reasoning | 57.7 % | independent | Artificial Analysis |
| SciCode | thinking | 39.1 % | independent | Artificial Analysis |
| SciCode | no reasoning | 36.7 % | independent | Artificial Analysis |
external indices
| Intelligence Index v4.1 | no reasoning | 13.7 points | independent | Artificial Analysis |
| Intelligence Index v4.1 | thinking | 13.5 points | independent | Artificial Analysis |
| AA Math Index | thinking | 89.7 points | independent | Artificial Analysis |
| AA Math Index | no reasoning | 49.7 points | independent | Artificial Analysis |
| ECI | 139.9 points | independent | Epoch AI Benchmarking Hub |
factuality
| MASK contamination=Potential contamination warning: This model was evaluated after the public release of MASK, allowing model builder access to the prompts and solutions. | 46.3 % 44.7–47.9 | independent | Scale Labs |
human preference
| Chinese (style control)older version arena=text · category=chinese · style_control=true | 1455.8 1439.9–1471.6 | community | LMArena Leaderboard Dataset | |
| Coding (style control)older version arena=text · category=coding · style_control=true | 1447.3 1435.7–1459.0 | community | LMArena Leaderboard Dataset | |
| Industry Software And It Services (style control)older version arena=text · category=industry_software_and_it_services · style_control=true | 1444.2 1435.2–1453.1 | community | LMArena Leaderboard Dataset | |
| Hard Prompts English (style control)older version arena=text · category=hard_prompts_english · style_control=true | 1436.1 1425.1–1447.1 | 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 | 1434.3 1421.9–1446.7 | community | LMArena Leaderboard Dataset | |
| Industry Medicine And Healthcare (style control)older version arena=text · category=industry_medicine_and_healthcare · style_control=true | 1433.4 1414.3–1452.6 | community | LMArena Leaderboard Dataset | |
| Hard Prompts (style control)older version arena=text · category=hard_prompts · style_control=true | 1432.4 1424.6–1440.2 | community | LMArena Leaderboard Dataset | |
| Expert (style control)older version arena=text · category=expert · style_control=true | 1427.8 1406.1–1449.6 | community | LMArena Leaderboard Dataset | |
| English (style control)older version arena=text · category=english · style_control=true | 1426.1 1417.9–1434.2 | community | LMArena Leaderboard Dataset | |
| Longer Query (style control)older version arena=text · category=longer_query · style_control=true | 1420.5 1410.0–1430.9 | community | LMArena Leaderboard Dataset | |
| Industry Legal And Government (style control)older version arena=text · category=industry_legal_and_government · style_control=true | 1418.5 1400.4–1436.7 | community | LMArena Leaderboard Dataset | |
| Overall (style control) arena=text · category=overall · style_control=true | 1417.4 1411.3–1423.4 | community | LMArena Leaderboard Dataset | |
| Math (style control)older version arena=text · category=math · style_control=true | 1413.8 1395.5–1432.2 | 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 | 1413.6 1401.7–1425.6 | community | LMArena Leaderboard Dataset | |
| Industry Mathematical (style control)older version arena=text · category=industry_mathematical · style_control=true | 1412.3 1392.6–1432.0 | community | LMArena Leaderboard Dataset | |
| Russian (style control)older version arena=text · category=russian · style_control=true | 1405.2 1386.8–1423.6 | community | LMArena Leaderboard Dataset | |
| Multi Turn (style control)older version arena=text · category=multi_turn · style_control=true | 1405.1 1393.2–1416.9 | community | LMArena Leaderboard Dataset | |
| Exclude Ties (style control)older version arena=text · category=exclude_ties · style_control=true | 1404.0 1395.7–1412.4 | community | LMArena Leaderboard Dataset | |
| Instruction Following (style control)older version arena=text · category=instruction_following · style_control=true | 1402.9 1393.1–1412.7 | community | LMArena Leaderboard Dataset | |
| Polish (style control)older version arena=text · category=polish · style_control=true | 1401.1 1381.9–1420.3 | community | LMArena Leaderboard Dataset | |
| Non English (style control)older version arena=text · category=non_english · style_control=true | 1398.8 1391.6–1406.0 | 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 | 1398.5 1388.2–1408.7 | community | LMArena Leaderboard Dataset | |
| Creative Writing (style control)older version arena=text · category=creative_writing · style_control=true | 1389.3 1375.8–1402.8 | 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 | 1377.0 1365.7–1388.4 | community | LMArena Leaderboard Dataset |
knowledge science
| GPQA Diamond implementation=artificial-analysis | thinking | 77.9 % | independent | Artificial Analysis |
| GPQA Diamond implementation=artificial-analysis | no reasoning | 73.5 % | independent | Artificial Analysis |
| Humanity's Last Exam implementation=artificial-analysis | thinking | 14.3 % | independent | Artificial Analysis |
| Humanity's Last Exam implementation=artificial-analysis | no reasoning | 6.7 % | independent | Artificial Analysis |
| MMLU-Pro implementation=artificial-analysis | thinking | 85.1 % | independent | Artificial Analysis |
| MMLU-Pro implementation=artificial-analysis | no reasoning | 83.3 % | independent | Artificial Analysis |
long context instruction
| AA-LCR | thinking | 56.7 % | independent | Artificial Analysis |
| AA-LCR | no reasoning | 47.0 % | independent | Artificial Analysis |
| IFBench | thinking | 41.5 % | independent | Artificial Analysis |
| IFBench | no reasoning | 37.8 % | independent | Artificial Analysis |
reasoning math
| AIME year=2025 · implementation=artificial-analysis | thinking | 89.7 % | independent | Artificial Analysis |
| AIME year=2025 · implementation=artificial-analysis | no reasoning | 49.7 % | independent | Artificial Analysis |
Agent + model results
systems, not bare-model scores
| agent + model Artificial Analysis harness + Deepseek V3 1 | Terminal-Bench Hard | 25.0 % | independent | Artificial Analysis |
| agent + model Artificial Analysis harness + Deepseek V3 1 | Terminal-Bench Hard | 24.2 % | independent | Artificial Analysis |
These scores measure the whole agent system (scaffold, tools, budgets) — they are never merged into the bare model’s numbers.