BenchLeader

Deepseek v2

Deepseek v2 is a DeepSeek open-weights model, released 7 May 2024. It is not yet ranked: 1 independent result so far, and the index needs at least five across two categories. Last measured 6 May 2024.stale: no new result in six months

Blended price
–
Output speed
–
First answer
–
Context
128k
Full answer
–
Cost per run
–
Released
7 May 2024
How it scores by categoryDashed line = average model (50). One step of 15 = one standard deviation.
  1. Composite36

Benchmark results

See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.

Reasoning

Benchmarknot statedSourceTrend
HellaSwagnot in index87.1%#5Epoch AI Benchmarking Hub
BIG-Bench Hardnot in index78.8%#8Epoch AI Benchmarking Hub
WinoGrandenot in index86.3%#5Epoch AI Benchmarking Hub
ARC (AI2 Challenge)not in index92.2%#4Epoch AI Benchmarking Hub
PIQAnot in index83.9%#8Epoch AI Benchmarking Hub

Knowledge

Benchmarknot statedSourceTrend
MMLUnot in index78.4%#27Epoch AI Benchmarking Hub
TriviaQAnot in index80.0%#15Epoch AI Benchmarking Hub

Composite

More from DeepSeek

New this month

See also

Questions people ask

Why does Deepseek v2 have no BenchLeader Index score?
The index needs at least five independent results across at least two categories; as of 25 Oct 2026 Deepseek v2 has 1 that count. Launch-post scores are never used, so it is ranked once independent publishers have measured it.
When was Deepseek v2 released?
DeepSeek released Deepseek v2 on 7 May 2024.

Data as of 25 Oct 2026. Compare with another model.

Cite as: BenchLeader, “Deepseek v2: benchmarks, pricing, speed and rank”, data as of 25 Oct 2026, https://www.benchleader.com/models/deepseek-v2.