DeepSeek V4.1-Flash-1017 发布:研发能力大幅提升,新增搜索API
版本概述
2026年10月17日,DeepSeek官方宣布将V4.1 Flash模型更新至V4.1-Flash-1017版本。此次更新同步上线了网页端、App与API,新模型在各项基准测试中均取得了显著提升。
核心更新
- 项目研发能力显著提升:新模型在代码生成、项目分析和复杂任务处理方面表现更优,能够更高效地完成研发相关工作。
- 成本更低:在保持高性能的同时,进一步优化了推理成本,为用户提供更具性价比的选择。
- 响应速度更快:完成任务的速度明显提升,用户体验更加流畅。
- 首次开放Search和Fetch API:新增网页搜索和内容抓取接口,价格均为5元/千次调用,正式向公众开放。
API定价
新增接口定价如下:
- Search接口:5元/千次调用
- Fetch接口:5元/千次调用
V4.1-Flash-1017模型本身的API定价保持不变,继续维持Flash系列的高性价比优势。
升级建议
建议所有DeepSeek用户尽快升级到最新版本,体验以下改进:
- 开发者可获得更强的代码辅助和项目分析能力
- 普通用户可享受更快的响应速度和更低的成本
- 开发者可通过新的Search和Fetch API构建更强大的AI应用
总结
DeepSeek V4.1-Flash-1017的发布标志着该系列模型在实用性和功能性上的重要进步。新增的搜索API为开发者打开了新的可能性,而模型本身的能力提升则让所有用户受益。作为国产大模型的代表,DeepSeek持续快速迭代,正在缩小与国际顶尖模型的差距。
DeepSeek V4.1-Flash-1017 Released: Major Gains in Development Capability, New Search API
Release Overview
On October 17, 2026, DeepSeek officially announced that the V4.1 Flash model had been updated to version V4.1-Flash-1017. The release went live simultaneously on the web app, the mobile app and the API, and the new model achieved significant gains across all benchmark tests.
Key Updates
- Significantly stronger project development capability: the new model performs better at code generation, project analysis and complex task handling, and can complete development-related work more efficiently.
- Lower cost: while maintaining high performance, inference costs were further optimized, giving users a more cost-effective option.
- Faster responses: tasks are completed noticeably faster, making the user experience smoother.
- Search and Fetch APIs opened for the first time: new web search and content fetching endpoints were added, both priced at ¥5 per thousand calls and now open to the public.
API Pricing
Pricing for the new endpoints is as follows:
- Search API: ¥5 per thousand calls
- Fetch API: ¥5 per thousand calls
Pricing for the V4.1-Flash-1017 model itself remains unchanged, keeping the Flash series' strong price-performance advantage.
Upgrade Recommendations
All DeepSeek users are advised to upgrade to the latest version to experience these improvements:
- Developers gain stronger code assistance and project analysis capabilities
- Everyday users enjoy faster response times and lower costs
- Developers can build more powerful AI applications with the new Search and Fetch APIs
Conclusion
The release of DeepSeek V4.1-Flash-1017 marks an important step forward for the series in terms of practicality and functionality. The new search APIs open up new possibilities for developers, while the model's own capability gains benefit all users. As a standard-bearer among China's domestic large models, DeepSeek keeps iterating rapidly and is closing the gap with the world's leading models.