NVIDIA Launches Nemotron 3.5 Super VL, a 120B Open Multimodal Model
NVIDIA has released Nemotron 3.5 Super VL, an open vision-language model designed for always-on enterprise agents. The model has 120 billion parameters. Source: marktechpost.com
115 articles
NVIDIA has released Nemotron 3.5 Super VL, an open vision-language model designed for always-on enterprise agents. The model has 120 billion parameters. Source: marktechpost.com
Multimodal agents usually produce freeform reasoning before every action, which can become lengthy and costly for small models with limited capacity. A new approach replaces open-ended reasoning with selection-based structured reasoning to build more efficient search agents. Source: machinelearning.apple.com
Always-on agents are handling longer and more complex enterprise workflows, where a single request may mean reading a document and cross-checking information across sources. NVIDIA's Nemotron 3.5 Super VL is aimed at powering such specialized agents with multimodal reasoning. Source: developer.nvidia.com
Muse Spark Open is an open-weight multimodal mixture-of-experts model built on an 816B-parameter text transformer backbone. SGLang and Miles support it from the day of release. Source: lmsys.org
Seedance 3.0 stretches single-clip generation to 30 seconds and expands its multimodal reference stack nearly sixfold. It targets creators and marketers building longer video content. Source: startupfortune.com
Cloudflare has released Clef-omni, a 30B decision model running on Workers AI that processes text, images, audio and video. It is aimed at content moderation and guardrail use cases. Source: tech-insider.org
Cloudflare launched Clef-omni on Oct. 9, a 30B-A3B multimodal decision model handling audio, video, image and text. Input is priced at $0.15 per million tokens. Source: shattered.io
Cloudflare has introduced Clef-omni, an open-weight decision model with 30B parameters (3B active) that can score audio, video, images and text in a single call. Source: marktechpost.com
AZ8 combines multimodal context, reusable creative Skills and intelligent model selection. The goal is to help creators produce high-quality video ads more efficiently. Source: tech.einnews.com
Google DeepMind's EmbeddingGemma 2 brings text, image, video and audio search to consumer devices. It needs less memory and depends less on cloud AI. Source: techrepublic.com
Google today released EmbeddingGemma 2, an open multimodal embedding model compact enough to run on a smartphone. It extends the original EmbeddingGemma line beyond text-only inputs. Source: siliconangle.com
EmbeddingGemma 2 is billed as the most capable model for on-device multimodal embeddings. It natively maps combinations of text, images, audio and video into a shared representation. Source: blog.google
A guide shows how to build multimodal search and RAG applications with EmbeddingGemma 2, a compact open model. It handles text, code, images, video and audio. Source: developers.googleblog.com
Google DeepMind's EmbeddingGemma 2 embeds text, code, images, video and audio in a single 740M-parameter model. It targets on-device RAG use cases. Source: marktechpost.com
Google DeepMind on Tuesday launched EmbeddingGemma 2, an open-weight AI model built to search and organize text, images, video and more. It is designed to run directly on devices. Source: tradingview.com
MixVio AI has launched a unified workspace that brings 26 AI generation routes together. Selected workflows accept up to 50 multimodal reference files and can produce 30-second 1080p video with natively synchronized audio. Source: manilatimes.net
The new tool combines images, video, text, audio, and reusable assets into a single connected visual workspace with AI-assisted production features. Source: einnews.com
Kling 4.0 marks a new step in AI-driven video creation, with focus on connecting multiple modalities into coherent, story-driven output. Details emerged from an announcement dated September 27, 2026, based in Albany, Alabama. Source: wboc.com
China presented what it calls the country's first multilingual, fully multimodal AI-powered input method for the Tibetan language. The system was unveiled on Sunday in Xining. Source: globaltimes.cn
An overview highlights five providers offering free access to LLM APIs, covering fast inference, multimodal capabilities, and agentic AI applications. The list is aimed at developers who want to experiment without paying for API usage. Source: kdnuggets.com