we've got more 𝚌𝚕𝚎𝚏 for you: • new 𝚌𝚕𝚎𝚏-𝚘𝚖𝚗𝚒 - takes audio, video, image, and text • 𝚌𝚕𝚎𝚏-𝚏𝚕𝚊𝚜𝚑 is now cheaper than jev • 𝚌𝚕𝚎𝚏 is now ~2x faster blog.cloudflare.com/clef-fas…

Introducing EmbeddingGemma 2, our new open embeddings model for on-device use cases! 👀 Code, image, video, audio, and text embeddings 🤏 Modular, going from 270m to 740m parameters 🪆 Matryoshka embeddings for storage efficiency 🤗 Apache 2 License

Meet EmbeddingGemma 2, our first natively multimodal open model for on-device embeddings. It expands beyond text to unify code, images, audio, and video in a shared space. 🧵

Ready to start building with EmbeddingGemma 2? Our Developer Guide walks you through embedding multimodal content across text, code, images, video, and audio using the sentence-transformers library and more! The workflow: 1️⃣ Load the model (via your preferred framework) 2️⃣ Embed text & code with task prompts 3️⃣ Embed images, video, audio & interleaved inputs 4️⃣ (Optional) Truncate dimensions with Matryoshka to save space

Introducing 𝗩𝗲𝗹𝗮 𝟮.𝟬: Our newest 𝗢𝗽𝗲𝗻 𝗙𝗼𝘂𝗻𝗱𝗮𝘁𝗶𝗼𝗻 𝗥𝗼𝘂𝘁𝗶𝗻𝗴 𝗠𝗼𝗱𝗲𝗹𝘀, bringing 𝘀𝗽𝗮𝗻-𝗹𝗲𝘃𝗲𝗹 𝗱𝗲𝗰𝗶𝘀𝗶𝗼𝗻𝘀 to System One. HF Collection 🤗: huggingface.co/collections/v… #OpenSource #AI #LLM #vLLM #HuggingFace

Introducing Decision 2.0: our newest state-of-the-art decision models, open in every size from 0.6B to 27B ⚡ HF collection 🤗: huggingface.co/collections/v… #OpenSource #Decision #Jev #LLM

Introducing Phonon-2: a new standard in speech recognition per byte. At just a tiny 164MB download, it is more accurate on average than OpenAI's Whisper large ( a model 10X its size) Transcribe an hour of audio in just 20 seconds on a MacBook Air! Open weights, CC-BY-4.0, today. @fermion_ai

We turned Qwen3.8-27B into a multimodal decision model. It beat Pokémon FireRed’s elite four and champion with sub-100 ms decisions from live game state. With SGLang’s native /v1/decisions, you can now turn LLMs and VLMs into classification and scoring models. We also added /v1/systemone so Jev-like open models can work with the TypeSafe SDK.

UPDATE: Kev-0.6B, 4B, and 8B are now available. Kev is a family of small open source Jev-like decision models you can train and run yourself. This new family is based on Qwen3 using the same LoRA + small pointer head technique as before, but scaled up. Out of domain, on data Kev never trained on: Kev-8B 79.6%, Jev 85.7%. • Drop-in TypeSafe System One API; their SDK works with one `base_url` change • Kev-4B serves on a 32 GB Mac in bf16: ~300 ms for five questions, ~40 ms on an H100 • Repeated documents hit a KV cache: 2-2.5x faster • Apache 2.0 License. Kev-4B trains in 40 minutes on one H100. Kev-8B in 83 minutes. Code, weights, evals: github.com/jaredpalmer/kev

Introducing SigLIP2: now trained with additional captioning and self-supervised losses! Stronger everywhere: - multilingual - cls. / ret. - localization - ocr - captioning / vqa Try it out, backward compatible! Models: github.com/google-research/b… Paper: arxiv.org/abs/2502.14786

Introducing GLiNER2.5-Decide, our new 340M parameter open weight, encoder-based decision model. GLiNER2.5-Decide is built for fast, deterministic classification. The model evaluates a set of user-defined typed questions and rules, and jointly decodes their answers, returning structured decisions with probability distributions and confidence scores. We evaluated the model’s performance on Fast Decisions, an unseen, internally generated classification suite based on 17 datasets testing real-world use cases across routing, triage, classification, sentiment, and content understanding. Measured against similar decision models, GLiNER2.5-Decide leads in 9 of the 17 datasets, achieving the highest average score: - GLiNER2.5-Decide: 60.1% - SemIf: 56.4% - JevK5: 57.5% - Laya: 46.6% This performance makes the model a strong fit for use cases like tool calling, model routing, browser and computer use, and LLM-as-a-judge. GLiNER2.5-Decide’s lightweight encoder architecture makes it easy to fine-tune the model for specific tasks, while being efficient enough to run locally on consumer-grade CPUs or in air-gapped environments, giving users greater control over where their data goes and where the model runs. To make building and experimenting with GLiNER2.5-Decide as easy as possible, we're also offering hosted inference. You can now use our API to run inference and fine-tune GLiNER models on specialized tasks right inside your own coding agent: agent.fastino.ai As with previous models, we’re also releasing the model weights on @huggingface under the Apache 2.0 license: huggingface.co/fastino/GLiNE…

When several people talk at once, a transcript can get messy fast. Our new Nemotron 3 Diarization model tracks who spoke when, even when voices overlap. It handles up to eight speakers, has 100M parameters, and is now available on @huggingface 🤗

The Jeff model, based on our GLiFormer, should be the most parameter-efficient model (a 500M-parameter model) for its performance score. And we created it right before the Jev release. Right now, imagine if we scale it 😉

Thanks for the model, we were able to port Laya to coreml with 99.5% of the ops on ANE + benchmarked too. it is now blazing fast with 3.7 ms per decision on an M5 Pro. Release: github.com/FluidInference/Fl… Models: huggingface.co/FluidInferenc…

This is NOT Jev. Open source. Runs on your laptop. Decides in ~27 ms, about 200× faster than waiting on a hosted LLM. Here it is playing Tetris by itself 👇 brainfunctioncollapse.com/la…