DeepSeek, the viral AI company, has released a new set of multimodal AI models that it claims can outperform OpenAI’s DALL-E 3.
The models, which are available for download from the AI dev platform Hugging Face, are a part of a new model family that DeepSeek is calling Janus Pro. They range in size from 1 billion parameters to 7 billion parameters. Parameters roughly correspond to a model’s problem-solving skills, and models with more parameters generally perform better than those with fewer parameters.
Janus Pro is under an MIT license, meaning it can be used commercially without restriction.
Janus Pro, which DeepSeek describes as a “novel autoregressive framework,” can both analyze and create new images. According to the company, on two AI evaluation benchmarks, GenEval and DPG-Bench, the largest Janus Pro model, Janus Pro 7B, beats DALL-E 3 as well as models such as PixArt-alpha, Emu3-Gen, and Stability AI‘s Stable Diffusion XL.
Some of those models are on the older side, granted. And Janus Pro can only analyze and generate small images — images 384 x 384 in resolution. But the Janus Pro family’s performance is impressive, considering the models’ compact sizes.
“Janus Pro surpasses previous unified model and matches or exceeds the performance of task-specific models,” DeepSeek writes in a post on Hugging Face. “The simplicity, high flexibility, and effectiveness of Janus Pro make it a strong candidate for next-generation unified multimodal models.”
DeepSeek, a Chinese AI lab funded largely by the quantitative trading firm High-Flyer Capital Management, broke into the mainstream consciousness this week after its chatbot app rose to the top of the Apple App Store charts. DeepSeek’s language models, which were trained using compute-efficient techniques, have led many Wall Street analysts — and technologists — to question whether the U.S. can maintain its lead in the AI race, and whether the demand for AI chips will sustain.
