Deepseek Coder V2 Lite Instruct
DeepSeek develops the Deepseek Coder V 2 Lite Instruct, a chat model exceling at code-related tasks such as code completion, generation, and review, as well as reasoning. Its Mixture-of-Experts (MoE) architecture and extended context length of 131,072 tokens enable substantial enhancements in coding and mathematical reasoning capabilities.
Input
Output
Context
131K
Max Output
164K
Parameters
15.7B
Technical Specifications
Model TypeChat
Context Window131,072 tokens
Max Output Tokens163,840 tokens
Parameters15.7B
Release DateJun 14, 2024
Training CutoffNov 1, 2023
Licenseother
Open SourceYes
Input Modalities
Text
Output Modalities
Text
Capabilities
Benchmarks
Artificial Analysis
5.3%
HLE
31.9%
GPQA
13.9%
SciCode
42.9%
MMLU Pro
15.8%
LiveCodeBench
3.1
Intelligence
Resources & Links
Estimated GPU Requirements for Deepseek Coder V2 Lite Instruct
Estimates based on INT8 quantization. Actual requirements vary by framework and configuration.
Related Tools
Data sourced from official provider APIs and documentation
Last updated: Jun 24, 2026
Start building with the right model.
Automatically route workloads to the right model for every task, every time.