Moonshot AI expands with new multimodal model and coding tool

China-based Moonshot AI, backed by investors including Alibaba and HongShan, today released Kimi K2.5, an open-source model capable of processing text, images, and video. Alongside the model, the company launched Kimi Code, an open-source coding tool that allows developers to use images and videos as inputs. The Kimi K2.5 model was trained on 15 trillion mixed visual and text tokens. According to the company, the model is capable of handling coding tasks and agent swarms, which are orchestrations where multiple agents work together. On the coding front, users can feed the model images or videos and ask it to generate a similar interface shown in those media files. In benchmarks, Kimi K2.5 matches or outperforms peers on several tasks. Specifically, Kimi K2.5 outperforms Gemini 3 Pro on SWE-Bench Verified (a software engineering benchmark) and scores higher than GPT 5.2 and Gemini 3 Pro on the SWE-Bench Multilingual benchmark. In video understanding, it outperforms GPT 5.2 on VideoMMMU (a benchmark for video understanding).

The launch of Kimi Code places Moonshot AI in competition with developer tools such as Anthropic’s Claude Code and Google’s Gemini. Developers can use Kimi Code through their terminals or integrate it with development software. Coding tools have become revenue drivers for AI developers. For context, Anthropic announced in November that Claude Code had reached $1 billion in annualized recurring revenue (ARR), and the tool reportedly added an additional $100 million to that figure by the end of 2025. Meanwhile, Moonshot AI’s competitor in China, DeepSeek, is set to release a new model with coding capabilities next month.

To support its expansion, Moonshot AI, which was founded by Yang Zhilin, is seeking to raise new capital. According to Bloomberg, the startup is seeking to raise a new funding round at a $5 billion valuation. This follows a progression of capital raises for the company:

• Series B Round: The company raised $1 billion in funding at a $2.5 billion valuation.
• Previous Round: Last month, the startup raised $500 million in funding at a $4.3 billion valuation.
• Current Target: Bloomberg reports that the company is seeking to raise a new round at a $5 billion valuation.

Why it matters

Moonshot AI is aggressively scaling its product suite to challenge major AI labs, using open-source releases to gain developer traction while simultaneously pushing for a higher valuation in the competitive Chinese market.