Huoshan Engine officially launched the new Doubao-Seed-Code programming model on November 11, and brought a surprising promotional event to developers, with the first month's subscription only costing 9.9 yuan. This move not only marks the latest breakthrough of Huoshan Engine in the AI programming field, but also provides developers with an unprecedented cost-effective choice.
The Doubao programming model is an artificial intelligence tool specifically optimized for Agentic programming tasks. It has set a new state-of-the-art (SOTA) record on the official SWE-Bench-Verified leaderboard, and is compatible with mainstream development environments such as the Anthropic API. According to Huoshan Engine, the Doubao programming model reduces the comprehensive usage cost by 62.7% compared to the industry average, making it the most affordable AI programming tool in China.
The model has a long context processing capability of 256K, which can easily handle complex code files and multi-module dependencies, greatly improving the efficiency of full-stack development. As the first programming model in China to support visual understanding capabilities, the Doubao programming model can generate code based on UI design drafts, screenshots, or hand-drawn sketches, and even automatically fix style issues and bugs, bringing revolutionary convenience to front-end developers.
Huoshan Engine has also specially launched the "Coding Plan" subscription package, offering more discounts for individual developers, with the first subscription costing only 9.9 yuan. This package supports Claude Code and is compatible with various mainstream tools such as veCLI, Cursor, and Cline, providing developers with stable computing power to ensure a smooth coding experience.
In addition, the Chinese version of TRAE has been officially integrated with the Doubao programming model, providing more comprehensive services for enterprise developers. TRAE, as the first AI-native IDE in China, is dedicated to providing intelligent Q&A, code auto-completion, and other AI programming capabilities. The enterprise version is now open for public testing, allowing enterprises to handle large code repositories in high-concurrency scenarios, and providing flexible model management and security control functions for enterprises.
