DeepSeek-V3-0324 released with comprehensive upgrades in reasoning, coding and Chinese writing
Original title:DeepSeek-V3 模型更新,各项能力全面进阶
AI summary
DeepSeek released the minor version upgrade DeepSeek-V3-0324, which scores higher than GPT-4.5 on mathematics and coding benchmarks and enhances front-end code generation, Chinese writing and web search capabilities. The model has about 660B parameters, the open-source version has a 128K context length, and the web version, App and API offer 64K context. The weights have been open-sourced under the MIT License, allowing distillation training of other models.
Key Takeaways
- DeepSeek released the minor version upgrade DeepSeek-V3-0324,which scores higher than GPT-4.
- 5 on mathematics and coding benchmarks and enhances front-end code generation,Chinese writing and…
- The model has about 660B parameters,the open-source version has a 128K context length,and the web…
- The weights have been open-sourced under the MIT License,allowing distillation training of other…
Quick Facts
- Published
- Updated
- Source
- 公众号:DeepSeek(深度求索)
- Category
- Model Releases
- AI score
- 74
推理能力提升至超越 GPT-4.5 的同时保持 MIT 开源许可,让低成本部署强推理模型的门槛进一步降低。
Related entities
Source & copyright
This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.
How to Cite
When citing this article, keep the title, date, and permalink, and credit Orange News as the source.
Orange News: DeepSeek-V3-0324 released with comprehensive upgrades in reasoning, coding and Chinese writing. March 25, 2025. https://new.nat6.net/en/news/cmsdu3h2e015lrofzs4upjdkf/