GenVR Research

Do you know that, Influencers and youtube content creators waste 30-40% of their time just reading through the comments on their social media handles. The situation is so tiresome that many times they need to employ 4-5 social media managers who make a summary of the actions on their posts. Presenting GenVR's (GenVR Research) first toolkit for social media management, "youtube comment analyzer" agent. The agent goes thoroughly over the youtube comment section of your video and your youtube channel, flags anything inappropriate, and makes a summary of bad and good comments. Effectively, it saves 60% of your bandwidth while managing your social handles and reduces the need for hiring expensive social media teams. A sample chat from v1 is attached here.

AI4Bhārat and IBM has recently conducted yet another benchmark study on Indic LLMs, comparing them on knowledge benchmark scores. This is published in following research paper:  https://lnkd.in/gUzrR_ki "MILU: AMulti-task Indic Language Understanding Benchmark" by Indian Institute of Technology, Madras, IBM at Nandan Nilekani center at AI4Bharat. We attach the comments about our model in the research paper and the model comparison scores as well. Where it comfortably outperforms Project Indus (by Tech Mahindra), or other models like Sarvam-1 and was the best finetuned Indic LLM.

Story Diffusion which can create exceptional storyboards is now live on GenVR Research's platform. With this, our model hub has now reached 75 models, running on over 500+ Nvidia GPUs.

Recraft V3 model (also known as Red Panda) is now live on GenVR's platform. The model is ranked #1 in the world as per ELO scores and had much higher win rate than Midjourney or Flux 1.1 pro. https://lnkd.in/gPZNBe9Q To try it out, login to app.genvrresearch.com and hover to Image generation section.

MBZUAI (Mohamed bin Zayed University of Artificial Intelligence) and Monojit Choudhury recently released new benchmarks for many Hindi LLMs in their Nanda release research paper and blog post:- https://lnkd.in/ggDe9Sxy https://lnkd.in/gWWDGVgh This also features Hindi LLMs from multiple companies like GenVR Research, Sarvam AI, Cohere, MBZUAI and AI4Bharat. Happy to share that our Aryabhatta Gemma series (released 6 months back) have stood the test of time and has yet again emerged among the top most Indic LLMs on benchmarks. Second best overall in current study done by MBZUAI. And ahead of Sarvam 0.5 B, Cohere's Aya model or AI4Bharat's work. Which has also been confirmed by numerous other benchmarks like Indic LLM Leaderboard (https://lnkd.in/gpqsiYBj ), Pariksha survey by Microsoft and now by MBZUAI. Although the recent study by MBZUAI lacks the recent models of GenVR (llama38GenZ) which has been the leading Hindi model in Pariksha survey Or Indic LLM leaderboard. We are happy to see how our other Gemma based models have performed. A big congrats to MBZUAI and Monojit Choudhury for the Nanda release.

We’re thrilled to introduce 𝗥𝗮𝗴𝗩𝗥—our innovative 3D asset search agent that’s set to transform how creators find the perfect assets for their projects. With over 𝟮𝟬 𝗺𝗶𝗹𝗹𝗶𝗼𝗻 3D assets gathered from leading online databases like Objaverse XL and Sketchfab, RagVR applies retrieval augmentation generation (RAG) on 3D assets and drastically reduces search time from 𝟭 𝗵𝗼𝘂𝗿 𝘁𝗼 𝗹𝗲𝘀𝘀 𝘁𝗵𝗮𝗻 𝟭𝟬 𝘀𝗲𝗰𝗼𝗻𝗱𝘀! ⏱️💡 This powerful tool is designed to empower artists, developers, and designers by making 3D asset discovery faster and more efficient than ever before. Ready to revolutionize your workflow? Check out 𝗥𝗮𝗴𝗩𝗥 and elevate your 3D projects! #GenVR #RagVR #3DAssets #Innovation #TechForCreators #DigitalArt #GameDev #3DModeling

Suno v3.5 and Midjourney v6 are now successfully integrated on GenVR Research's web application. This comes after we successfully integrated popular models like Ideogram V2, Kling Pro, CogVideoX, Stable Diffusion 3.5, Luma Dream Machine. This will give higher flexibility to our subscribers to create art work, music, video, images, etc. Now one can literally do anything on GenVR's web application. Be it generating high quality music, image generation, video generation, talking to 1000+ AI personas, makeup transformation, panorama generation, image makeovers, etc. etc. etc. A collection of 100+ features and models.

We just added a new section on our web application called Video Gen. Even though, we don't work on finetuning or creating video gen models, based on popular demand, we have decided to provide all our subscribers access to best video gen models (like CogVideoX, Luma Dream Machine and Kling Pro).

Happy to share that GenVR Research's web application now supports Stable Diffusion 3.5 (released today) and Ideogram V2.

GenVR Research's makeup try on application is now live. Originally designed as a part of JioGenNext project, we are rolling out the access to all our community members. Attached are few images of a single subject undergoing eyebrow and lip color change. To try it out, login to app.genvrresearch.com and hover over to Image makeover section.