Liquid AI's LFM2.5-24B Runs 184 Tokens per Second for Local Agents
Liquid AI's new 24B mixture-of-experts activates only 2B parameters per token, delivers 184 tokens/sec on an M5 Max, and ships in GGUF for llama.cpp.
20 stories
Liquid AI's new 24B mixture-of-experts activates only 2B parameters per token, delivers 184 tokens/sec on an M5 Max, and ships in GGUF for llama.cpp.
Bria's 8B Fibo 2.0 unifies image generation and editing, tops open-weights leaderboards, and ships trained only on licensed data.
A trending open-source SLAM system fuses 3D Gaussian Splatting with semantic segmentation to deliver object-aware dense reconstruction in real time.
A Rust-based filesystem gives each AI agent a disposable, zero-copy fork of a shared base, enabling cheap branching, rollback, and parallel rollouts.
A new reinforcement pre-training recipe pushes rewards into the hidden iterations of looped transformers, improving accuracy and cutting compute per token.
Geo-sleuth is an open-source agent skill that locates photos to within meters using OpenStreetMap, elevation skylines, satellite tiles and street view, with no landmarks required.
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 new benchmark exposes how even Sora 2 and Veo 3.1 fail at generating physically plausible sounds, scoring under 0.46 on audio-physics tests.
A weight-level edit strips refusals from DeepSeek V4.1 Flash while keeping FP8 precision, 1M context, vision, and tools intact.
A unified CuTe DSL decode path on GB300 pushes Kimi K3 to 842 tokens/s at batch 1, up to 2x faster than stock vLLM.
A new homebrew project ports llama.cpp to jailbroken PlayStation 5 consoles, aiming to run Qwen 3.8 27B natively on Sony hardware.
A training-free token compression method keeps Video LLMs accurate even when 99% of visual tokens are thrown away, with a 12.5x prefill speedup.
A new study from Robocurve measures how fast general-purpose language models are learning to drive robot arms, and the curve is steep.
DreamCan skips dense video prediction and plans with a handful of carefully chosen keyframes, hitting 91.2% success on LIBERO-Long with a 0.1B policy.
A new open dataset of 8,800 synthetic doctor-patient conversations with 1,329 hours of audio tests whether audio language models can summarize long clinical visits.
A new MIT-licensed repo recreates ChatGPT's Intelligent UI, streaming interactive maps, charts, forms, and mini apps instead of plain text replies.
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.
Datalab's new 1M-parameter Indra OCR reads 49 pages per second on CPU, outpacing Tesseract and PaddleOCR while keeping layout and reading order intact.
ScrapeCreators shipped a plugin bundle of 13 agent skills that lets coding assistants run full social media research workflows across TikTok, Instagram, YouTube, Reddit, X, LinkedIn and more.
A new benchmark of 1,246 questions over 13,956 AI papers exposes how badly current LLM agents handle real literature-review workflows.