ModelDeepSeek Coder V2

DeepSeek-Coder-V2

A code-focused large language model based on a Mixture-of-Experts architecture, released by DeepSeek. The full model reaches 236 billion parameters, and a lightweight distilled version is also available for local deployment.

Timeline (last 90 days)

Sep 1

The performance of Qwen2.5-Coder and DeepSeek-Coder-V2 on code benchmarks such as HumanEval and MBPP is already very close to that of top closed-source models such as Claude 3.5 Sonnet.

Unverified50%

All Facts (1)

Unverified

The performance of Qwen2.5-Coder and DeepSeek-Coder-V2 on code benchmarks such as HumanEval and MBPP is already very close to that of top closed-source models such as Claude 3.5 Sonnet.

50%

Source Articles

Which LLM Writes the Most Efficient C/Rust Code? We Tested 4 Models on Low-Level OptimizationOct 11 RTX 5060 Ti 16GB Local LLM Selection Guide: Ollama Model Recommendations and BenchmarksSep 24