Inspired by fofr's recent sinister robot prompting. I updated the replicate.com/replicate-prun… endpoint on Replicate to optimise prompts by translating them to chinese to automatically get better results. This is generated in 15s at the price of an image model (2.5c) > a

Really cool community contribution that showcases how to optimise models through Pruna and deploy them FastAPI!

Qwen image with Replicate lora, runs in ~6s using the super fast pruna-optimised model Pass your trained weights as a `lora_weights` input to replicate.com/qwen/qwen-image Example: replicate.com/p/kpqmcywygnrm…

.Qwen

Qwen Image Edit. 3 seconds, $0.03. replicate.com/qwen/qwen-imag… Pruna AI do the impossible. > Make the text 3D and floating on a city street

20x Faster Qwen-Image-Edit optimised in under 30 mins.

We are really happy to collaborate with Replicate to bring you a 3s per generation Qwen-Image-Edit!

So cool!

We updated Qwen-Image-Edit to have better quality and consistency and by popular demand aspect ratios

🚀 𝗠𝗮𝘀𝘀𝗶𝘃𝗲 𝘂𝗽𝗴𝗿𝗮𝗱𝗲 𝗳𝗼𝗿 𝗤𝘄𝗲𝗻-𝗜𝗺𝗮𝗴𝗲! We've just dropped new speed, memory, and feature improvements on Replicate that are going to change your AI image generation workflow: • 𝗦𝗽𝗲𝗲𝗱 𝗯𝗼𝗼𝘀𝘁- Now generates in just 3 seconds! • 𝗠𝗲𝗺𝗼𝗿𝘆

Love seeing Replicate users value our fast and cheap endpoints when building 💜

It’s not every day that you 𝐜𝐮𝐭 𝐀𝐈 𝐢𝐧𝐟𝐞𝐫𝐞𝐧𝐜𝐞 𝐭𝐢𝐦𝐞 𝐛𝐲 𝟓𝟎%! Our new blog post tells the story of how we did it in just 2 days, without losing quality, using torch compile, and with the help of our partners at Pruna AI. Read the full story on our blog