Home/Models/DeepSeek-V2 (MoE-236B, May 2024)
DeepSeek logo

DeepSeek-V2 (MoE-236B, May 2024)

by DeepSeek · Released Jan 2024

Open Source
84.4
avg score
Rank #16
Compare
Better than 93% of all models
Context
N/A
Input $/1M
TBD
Output $/1M
TBD
Type
text
License
Open Source
Benchmarks
7 tested
Data updated today
About

Tested on 7 benchmarks with 76.5% average. Top scores: ARC AI2 (89.6%), HellaSwag (82.8%), TriviaQA (80.0%).

Capabilities
reasoning
71.7
#22 globally
knowledge
77.3
#6 globally
Benchmark Scores
Compare All
Tested on 7 benchmarks · Ranked across 2 categories
Score Distribution (all 233 models)
0255075100
▲ You are here
BBH

BIG-Bench Hard. 23 challenging tasks from BIG-Bench where prior language models fell below average human performance.

71.7
ARC AI2

AI2 Reasoning Challenge. Grade-school science questions requiring multi-step reasoning. Easy and Challenge sets test different difficulty levels.

89.6
HellaSwag

Sentence completion requiring commonsense reasoning about physical and social situations. Tests real-world understanding.

82.8
TriviaQA

Trivia questions sourced from trivia enthusiasts and quiz websites. Tests breadth of general knowledge.

80.0
Excellent (85+) Good (70-85) Average (50-70) Below (<50)
Links
Documentation
Community
BenchGecko API
deepseek-v2-moe-236b-may-2024
Specifications
  • Typetext
  • ContextN/A
  • ReleasedJan 2024
  • LicenseOpen Source
  • Statusbenchmark-only
Available On
DeepSeek logoDeepSeekTBD
Share & Export
Tweet
DeepSeek-V2 (MoE-236B, May 2024) is an open-source text AI model by DeepSeek, released in January 2024. It has an average benchmark score of 84.4.