LFM2.5
PulseAugur coverage of LFM2.5 — every cluster mentioning LFM2.5 across labs, papers, and developer communities, ranked by signal.
2 day(s) with sentiment data
-
Liquid AI launches RIPL engine for sub-50ms LFM inference
Liquid AI has introduced RIPL, a new real-time inference engine designed to serve its Liquid Foundation Models (LFMs) with very low latency. The engine aims to enable the use of generative models in applications that pr…
-
Open-source app Private Mind adds on-device web search for LLMs
The open-source app Private Mind has integrated web search capabilities for its on-device LLM functionality on iOS and Android. This feature utilizes a 1.7B parameter model, Qwen 3, to process and cite information from …
-
Draft model for LFM2.5 released, offering up to 3.2x speed increase
A draft model designed for LFM2.5 has been released for free, offering up to a 3.2x speed improvement. This development aims to enhance the performance of AI systems utilizing the LFM2.5 framework.
-
Liquid AI boosts LFM2.5 model speed up to 3.18x with DSpark speculative decoding
Liquid AI has released DSpark draft models for its LFM2.5 series, which enhance decoding speed by up to 3.18x without altering output quality. These models utilize speculative decoding, where a smaller draft model propo…
-
Liquid AI releases QAD versions of LFM2.5 models for efficient edge deployment
Liquid AI has released a Quantization-Aware Distillation (QAD) version of its LFM2.5 series of small AI models. This new version is designed to reduce memory usage while minimizing performance degradation, making the mo…
-
32 local LLMs tested head-to-head; most show similar performance
A comprehensive head-to-head comparison of 32 local large language models (LLMs) on a fact-extraction corpus revealed that most models performed similarly. The study, which utilized paired bootstrap testing on consumer-…
-
Hugging Face integrates GGML and llama.cpp for local AI development
Hugging Face has announced the integration of GGML and llama.cpp, two key projects for running large language models locally on consumer hardware. This move aims to ensure the long-term development and accessibility of …
-
Masked Diffusion Language Models outperform AR models for agentic RL
A new research paper introduces Masked Diffusion Language Models (MDLMs) as a superior alternative to autoregressive (AR) models for text-based world modeling in agentic reinforcement learning. MDLMs demonstrate enhance…
-
20 small LLMs benchmarked on 6GB GPU for practical use
A user benchmarked 20 small language models on a 6GB RTX 4050 GPU to assess their practical utility for overnight tasks like file organization and log triage. The evaluation focused on qualitative tests and performance …
-
Ollama releases new open-source models minimax-m3 and lfm2.5
Ollama has released two new open-source models, minimax-m3 and lfm2.5. These models are available for download through the Ollama library, providing users with new options for running AI locally. The releases suggest co…
-
Liquid AI launches Japanese text and audio LLMs
Liquid AI has released two new Japanese-language models: LFM2.5-1.2B-JP-202606, a text-based chat model that shows improved performance in knowledge, instruction following, and coding, and LFM2.5-Audio-1.5B-JP, a multim…