Liquid AI released LFM2.5-24B-A2B on Oct. 13, an open-weight reasoning model with 24B total parameters and roughly 2B active per token (Liquid AI elsewhere cites 2.45B), aimed at running agents locally on laptops and PCs. It is the largest and final member of the LFM2.5 family, with the context window extended from 32K to 131K tokens, a vocabulary doubled from 65K to 128K for better non-Latin tokenization, pretraining scaled from about 17T to 33T tokens, and a switch to reasoning-only output with an explicit chain of thought. Liquid AI says it matches the larger Qwen3.6-35B-A3B on five of eight agentic and tool-use benchmarks (for example 48.09 versus 38.19 on LiveClaw), though those benchmarks and the speed figures (184 tokens/s on an Apple M5 Max) are vendor-run. Base and post-trained weights are on Hugging Face with day-one support in llama.cpp, MLX, vLLM, SGLang and ONNX.
Rapid-MLX is an Apache-2.0 local LLM inference server built on MLX for Apple Silicon, exposing both OpenAI (/v1/chat/completions, /v1/responses) and Anthropic Messages (/v1/messages) APIs and focusing on reliable tool calling for coding agents. The project claims up to 4× faster decoding than Apple's mlx-lm on identical weights: a 1.50× median across 18 streaming tasks with Qwen3.5-9B 4-bit on an M4 Pro Mac mini, and up to 4.27× on a whole-file code edit (46.4 to 198.4 tok/s), plus 3.0× Ollama's aggregate decode at 8 concurrent streams on Qwen3.6-35B-A3B. All figures are the project's own measurements, and much of its comparison table is drawn from other projects' own documentation.
Ollama v0.40.0 makes MLX the default runtime for supported model architectures on Apple Silicon. The release also adds MLX support for decision models and the EmbeddingGemma 2 embedding model, with additional model families listed in the release notes.
What problem it solves
People often want an assistant that can participate in ongoing conversations, remember personal context, and use tools without requiring a cloud AI service. Jarvis addresses this with local-first voice and text interaction, personal memory, optional connected tools, and a setup process for locally controlled models.
Key features
Conversation-aware voice assistance — Jarvis m…
What problem it solves
Serving language models requires reading checkpoints, tokenizing requests, and running accelerated kernels while managing a local model server. TensorFold addresses these requirements with a native binary that can run on Apple Silicon and NVIDIA GPUs without requiring Python or MLX for serving.
Key features
Native Metal and CUDA inference — The Zig-based engine executes mode…
What problem it solves
Time-series forecasting requires models that can forecast univariate and multivariate series across different context lengths. Forecasters also need to incorporate optional covariates and produce quantile forecasts without requiring per-task tuning.
Key features
Univariate and multivariate forecasting — Forecast individual time series or multiple target variates from NumPy a…