Recently, in the benchmark test of the crowdsourcing analysis platform Artificial Analysis, a mysterious AI image generation model code-named red_panda suddenly emerged, and its performance shocked the industry. It surpassed industry giants such as Midjourney, Black Forest Labs and OpenAI in multiple key indicators, attracting widespread attention. The editor of Downcodes will give you an in-depth understanding of the extraordinary strength of red_panda and its impact on the field of AI image generation.
Recently, a mysterious AI image generation model code-named red_panda achieved amazing results in the benchmark test of the crowdsourcing analysis platform Artificial Analysis, surpassing products from industry leaders such as Midjourney, Black Forest Labs and OpenAI by a significant advantage.
According to data from the Artificial Analysis platform, red_panda leads the text-to-image test with a score of 1244 points, surpassing the second-ranked Flux1.1Pro of Black Forest Labs by about 40 Elo points. It is worth mentioning that this scoring system is derived from the chess player strength evaluation system and is now used to compare the performance of AI models. Judging from the actual battle data, red_panda's winning rate is 9% higher than Flux, showing a clear technical advantage.
In terms of performance, red_panda is equally impressive. This model takes only 7 seconds on average to complete the generation of an image, which is more than a hundred times more efficient than OpenAI's DALL-E3. Through comparison of actual test samples, it was found that the images generated by red_panda have a high degree of realism, almost completely getting rid of the unnatural traces common in traditional AI works, and even surpassing the Flux model, which is famous for its realism, in terms of realism. At the same time, the model also performs well in understanding and executing text prompts.
Currently, red_panda is still in the closed beta stage and has not yet been opened to the public. Regarding the development team behind it, there are many speculations in the industry: some people think that this may be the incarnation of Midjourney-V7, and some analysts point out that this may be a new work from a Chinese technology company. However, judging from the technical characteristics, the model is more likely to be a real Lora fine-tuning model based on SD3.5large turbo training, rather than coming from a domestic mainstream technology company.
With the rapid development of AI image generation technology, the emergence of red_panda has once again promoted the improvement of industry standards. Where this mysterious new player comes from and how its leading technology will affect the development trend of the entire industry still deserves continued attention.
Artificial Analysis address: https://artificialanalysis.ai/text-to-image/arena
The emergence of red_panda has undoubtedly injected a boost into the field of AI image generation. Its efficient generation speed and high-quality image effects are exciting. In the future, we will continue to pay attention to the subsequent development of red_panda and look forward to it bringing us more surprises.