Overall score
52.6
Rank #50 · A confidence
DeepSeek · preview
DeepSeek's official V4 API model with optional thinking, tool use and open model weights.
Overall score
52.6
Rank #50 · A confidence
Capability
43.0
23 eligible benchmark families
Coverage
81%
Confidence A
Benchmarks scored
27
31 raw result rows
Model specification
Identity, configuration and price are preserved as dated source records.
Category profile
A category remains Not evaluated until an eligible source-qualified result is attached.
Benchmark scores
Best published score per benchmark. Open any row for the full per-benchmark leaderboard.
| Benchmark | Category | Score | Date |
|---|---|---|---|
| τ²-Bench TelecomRanking weighted | agents | 96.1988% | 2026-07-15 |
| GPQA DiamondRanking weighted | reasoning | 88.79% | 2026-07-15 |
| BrowseCompRanking weighted | agents | 83.4% | 2026-07-20 |
| MMLU-ProRanking weighted | research | 82.9% | 2026-07-15 |
| SWE-bench VerifiedRanking weighted | coding | 73.6% | 2026-07-15 |
| MCP AtlasRanking weighted | agents | 69.4% | 2026-07-15 |
| AA Long Context ReasoningRanking weighted | research | 66.3333% | 2026-07-15 |
| Terminal-BenchRanking weighted | agents | 64.04% | 2026-07-15 |
| BenchLM Coding priorRanking weighted | coding | 59.28% | 2026-07-15 |
| LiveCodeBenchRanking weighted | coding | 56.8% | 2026-07-15 |
| BenchLM Agentic priorRanking weighted | agents | 52.21% | 2026-07-15 |
| SWE-Bench ProRanking weighted | coding | 52.1% | 2026-07-15 |
| BenchLM Knowledge priorRanking weighted | research | 48.4% | 2026-07-15 |
| ToolathlonRanking weighted | agents | 46.3% | 2026-07-15 |
| Terminal-Bench HardRanking weighted | agents | 46.2121% | 2026-07-15 |
| SimpleQARanking weighted | research | 45% | 2026-07-15 |
| EnterpriseOps-Gym-AARanking weighted | agents | 40.4357% | 2026-07-15 |
| GDPval-AA v2Ranking weighted | agents | 40.3605% | 2026-07-15 |
| ITBench-AARanking weighted | agents | 38.3239% | 2026-07-15 |
| Humanity's Last ExamRanking weighted | reasoning | 35.87% | 2026-07-15 |
| HMMT Feb 2026Ranking weighted | mathematics | 31.7% | 2026-07-15 |
| τ³-BankingRanking weighted | agents | 25.7732% | 2026-07-15 |
| BenchLM Math priorRanking weighted | mathematics | 24.8% | 2026-07-15 |
| APEX-Agents-AARanking weighted | agents | 24.2625% | 2026-07-15 |
| CritPtRanking weighted | reasoning | 12.8571% | 2026-07-15 |
| Harvey LAB-AARanking weighted | agents | 3.3333% | 2026-07-15 |
| AA-OmniscienceRanking weighted | research | -10.0167 index | 2026-07-15 |
Showing the best of 31 raw result rows across 27 benchmark families.
Price and efficiency
Operational values are shown only when the source publishes a comparable measurement or a valid derivation.
Verification summary
News and X posts are structurally excluded from this model's evidence graph.
Missing publication dimensions: value. The last source check for this exact model was 2026-07-14.
Related models
Exact variants remain separate even when they share a provider or family.
DeepSeek V4 · preview
DeepSeek V3.2DeepSeek V3 · stable
DeepSeek V3.1 TerminusDeepSeek V3 · stable