Qwen Ships Qwen3-VL 8B Locally, Seeing Images and Video in 5 GB
Alibaba's Qwen team shipped GGUF weights for Qwen3-VL-8B-Instruct, letting the multimodal model run locally through llama.cpp, Ollama, and LM Studio.
20 stories
Alibaba's Qwen team shipped GGUF weights for Qwen3-VL-8B-Instruct, letting the multimodal model run locally through llama.cpp, Ollama, and LM Studio.
A weight-level edit strips refusals from DeepSeek V4.1 Flash while keeping FP8 precision, 1M context, vision, and tools intact.
Apple's DFN5B-CLIP-ViT-H-14-378 keeps climbing Hugging Face charts, hitting 84.2% zero-shot ImageNet accuracy by rethinking how training data is filtered.
Comfy-Org repackaged the FLUX.2 klein 4B text encoder and VAE for drop-in ComfyUI use, pushing the sub-second open model past 800k downloads.
Unsloth's reupload of Meta's 1B instruction-tuned Llama ships GGUF and 4-bit quantizations tuned for 2x faster finetuning with 70% less memory.
MiniMax dropped the open weights for its 428B parameter multimodal MoE model, with 23B active per token, 1M context, and strong agentic benchmarks.
A community mashup stacks 40 persona modes on top of uncensored Qwen3.8-27B weights, runs on a single 24GB GPU, and ships ready for llama.cpp and LM Studio.
ReadyArt drops a 31B uncensored Gemma-based roleplay model in GGUF format, built for explicit fiction, horror, and violence on local hardware.
NVIDIA's new 4B model reuses a shared block of 20 layers up to four times, trading compute for accuracy without retraining.
Fish Audio's new 35B mixture-of-experts TTS model takes bracketed stage directions, tops open-weight leaderboards, and ships free under a revenue cap.
Ant Group's InclusionAI opens weights for a 560B hybrid-linear MoE model with 25B active parameters, 1M context, and MIT licensing.
Comfy-Org repackaged the OpenMOSS 5B diffusion transformer for ComfyUI, bringing JSON layout prompts and strong text rendering to local workflows under Apache 2.0.
A community-fine-tuned Turkish speech recognition model built on Meta's XLS-R backbone has quietly crossed 893K downloads on Hugging Face.
A 27B uncensored model with multi-token prediction runs at 50 tok/s on an RTX 3060, fits on 8GB cards, and refuses nothing.
Perplexity's new 27B and 4B decision models return calibrated probabilities instead of text, answer multiple questions in one pass, and cost two cents per million tokens.
QuantTrio shipped a 4-bit AWQ quantization of Qwen3.6-35B-A3B that compresses the MoE down to roughly 24GiB for vLLM deployment.
OpenBMB's MiniCPM-V 4.7 packs 35B parameters into a mixture-of-experts vision model that runs on an iPhone at 13+ tokens per second.
A 315M-parameter Japanese embedding model built on ModernBERT-Ja tops JMTEB benchmarks while handling 8K context, outperforming OpenAI and billion-parameter rivals.
A 31B Gemma 4 fine-tune turns open language models into fast, calibrated decision engines for routing, extraction and classification in Polish and English.
Paradee-8M v1.1 is an 8.5MB text-to-speech model distilled from Kokoro-82M that runs 62x faster than real time on a single CPU thread.