#Multimodal

AISGLang says it worked with NVIDIA to optimize attention, MoE, and speculative verification kernels for Kimi K3 inference on early-access Rubin hardware. It reports up to 20% faster FP8 MLA at batch 1 with 128K context, 20% faster KDA verification with bitwise-identical output, and a 5.9% end-to-end speedup from MoE tail fusion that removes 276 kernel launches per decode step. The post also says SGLang powers Miles' end-to-end RL training on Rubin, including agentic RL with 64 concurrent sandboxes on the Vera CPU.

AISyren Video, a new agentic video tool, learns preferred graphics, motion, and editing rhythm from a user's library and generates new videos from a prompt. Users refine the results through chat, and the tool is free to try in a browser or through Claude MCP, per the company's announcement. The post's author says the education sector is exploring it.

AIThe vLLM Semantic Router team has released Decision 2.0, which answers multiple questions about one input in a single forward pass and outputs per-option probabilities. The post presents this as useful for routing and classification. A quoted post from Xunzhuo Liu says Decision 2.0 includes six open decision models ranging from 0.6B to 27B parameters, each topping same-size open models on the Jev Decision Index 0.3.

AIllama.cpp v0.6.0 adds support for multimodal Jev-like models and a performance upgrade for Metal, according to Merve Noyan. The post says models can be served with a single command, `llama serve -hf ggml-org/Clef-Flash-GGUF`, and points to a trending list of decision models on Hugging Face.

AIvLLM-Omni now supports Kandinsky 6.0 Video from launch day, with inference ready at release. Kandinsky 6.0 Video generates 5-second clips with synchronized audio and lip-sync from text or image inputs. Kandinsky's code and checkpoints are released under the MIT license, with Lite (3B) and Pro (29B) variants.

AIThe llama.cpp v0.6.0 release adds Clef support for text and vision, along with high-quality support for Qwen3.8-Flash-Next. It also brings a major Metal performance improvement and a new llama_batch_ext API, and the project website at llama.app has been refreshed.

AIAlephAlpha has released Kolibri, a German-English model with 78B total parameters, 3.46B active, and context up to 1M tokens. The weights are available under Apache 2.0 for running on users' own hardware. Cohere's Aidan Gomez congratulated the team on the model and its detailed technical report.

AISGLang demonstrated Qwen3.8-27B as a multimodal decision model that beat Pokémon FireRed's Elite Four and champion with sub-100 ms decisions from live game state. The company says its native /v1/decisions API lets LLMs and VLMs be used for classification and scoring. It also announced /v1/systemone for running Jev-like open models with the TypeSafe SDK.

AIPika has launched Camera Director, an app that reframes any shot to show it from different camera angles while preserving the original content. A quoted post praises its quality and says it is useful for real-world work.

AIPika has launched a new AI creative platform offering more than 40 apps covering needs from finishing edits to full video ads. The company describes the product as built for creatives and designed to meet exacting output standards.

AIReplicate says it is powering Griffin from Tavus on its platform. Tavus describes Griffin as the first model to pass the video Turing test, with 48% of live conversation participants believing it was a real human. The model ranks first on NVIDIA's full-duplex AI video benchmark.

AIIdeogram 4.5 was built for precise, targeted editing but also performs very well across general editing tasks. Design Arena ranks it 15th in Image Editing with an Elo of 1250, placing it in the same performance band as MAI-Image-2.6 and Gemini 3 Pro Image Preview. It is especially strong at typography edits, such as modifying text in infographics.

AIOpenBMB announced that MiniCPM-o 4.5 is now supported in SGLang Omni v0.1.7, giving developers more flexibility to run and build with the model. The background release notes add that MiniCPM-o 4.5 brings multimodal input and speech output to the runtime. MiniCPM-o and MiniMax-Music3 also gained Intel XPU support in the same release.

AIA developer connected a Raspberry Pi Zero W2 with a webcam and ambient microphone to MiniCPM-o 4.5, an open-source model running locally on a PC. The setup streams audio and video in real time with full-duplex interaction and no cloud APIs.

AIOpenBMB says Dubedo is the kind of product it hoped VoxCPM2 would enable, citing speaker-aware cloning, multilingual generation, and an editing studio. The post praises @dubedostudio's work, while the background post describes Dubedo as dubbing into 30 languages with per-speaker voice cloning and a beta open for trial.

AIIntern-S2-397B, a model built for long-horizon scientific research, now has Day-0 support in vLLM. The model brings multimodal, reasoning, coding, and scientific agent capabilities, and vLLM has published a run recipe for it.