
The company says its updated model is available online and through its app and API, while search tools cost $0.75 per 1,000 calls.
DeepSeek released its updated V4.1-Flash model and opened its search and retrieval APIs to developers. The company said the model, named DeepSeek-V4.1-Flash-1017, is available on its website, mobile app and through an API.
DeepSeek described the update as a performance and efficiency improvement without higher costs. The model scored 77.0 on the DeepSWE v1.1 test and 60.5 on Automation-Bench V1.0.6, according to results shared by the company.
Each search and retrieval API endpoint costs $0.75 per 1,000 calls, DeepSeek said. The company also expects the model to improve performance on larger development projects, including building games and creating 3D models from scratch.
This article was produced with the help of AI technology. Based on reporting by the outlets listed above.