Artificial Analysis, in partnership with Liquid AI, has launched a benchmark suite for pocket-scale AI models that fit within 8 GB of memory after quantization, including KV cache at 8K context, measu…
Liquid AI released Pipette, an open-source benchmarking platform for foundation models on edge devices, developed with Artificial Analysis as an independent methodology validator. Pipette measures on-…
Liquid AI released DSpark draft model checkpoints for three LFM2.5 models, enabling speculative decoding that speeds up inference by up to 3.18x on GPUs and 2.87x on-device without changing output qua…
Liquid AI, an MIT spinout, released four updated LFM2.5 checkpoints on August 19, trained with quantization-aware distillation (QAD) to retain roughly 97% of their BF16 average benchmark performance w…
Liquid AI released DSpark draft model checkpoints for three LFM2.5 models, delivering up to 3.18x faster decoding on an H100 and up to 2.87x on an M4 Max MacBook Pro without changing model outputs. Th…
Liquid AI released DSpark draft model checkpoints for its LFM2.5 family, claiming up to 3.18x throughput improvement on GPU and up to 2.87x on-device, with day-one support for llama.cpp and SGLang. Th…
Liquid AI released Q4_0 checkpoints for its LFM2.5 language models trained with quantization-aware distillation (QAD), recovering 96.5% to 97.4% of the accuracy lost to quantization across four models…
Liquid AI, an AI company, released an open-source byte-pair encoding (BPE) tokenizer trainer called toktoktok on GitHub, built autonomously by two coding agents using Claude Opus 4.5 and Codex with GP…
Liquid AI released LFM2.5-VL-3B, a 3.1B-parameter vision-language model for on-device deployment, averaging 69.4 across 28 vision benchmarks, matching InternVL-3.5-4B and 0.7 points behind Qwen3.5-4B.…
Liquid AI released LFM2.5-VL-3B, a 3.1B parameter vision-language model for edge devices, claiming it leads its size class on real-world image tasks while supporting screen/UI understanding, grounding…
Liquid AI released LFM2.5-2.6B, an on-device agentic model with 2.69B total parameters, a 131,072-token context window, and a 128,000-token vocabulary, pre-trained on approximately 34 trillion tokens.…
MacPaw announced a long-term strategic partnership with Liquid AI to co-develop an on-device AI technology stack for macOS, starting with its Eney AI assistant. The collaboration will adapt Liquid AI'…
MacPaw, a Ukraine-based app developer, has partnered with Liquid AI to power its products with locally hosted AI models, including an on-device inference system called Elix and a local memory system f…
Liquid AI's LFM2.5-2.6B, a 2.6-billion-parameter hybrid Mamba-Transformer model, can be deployed as a local AI agent on a single consumer GPU with 8–12 GB VRAM, according to a hands-on walkthrough. Th…
Liquid AI released LFM2.5-2.6B, a 2.6B dense model for on-device agentic workloads with a 128K context window and native tool calling, designed to run on edge devices and integrate with agent harnesse…
Liquid AI released LFM2.5-2.6B, a 2.6-billion-parameter on-device agentic model that outperforms models up to 4x larger on tool use and instruction following, achieving 220 tokens per second on an App…
Liquid AI released two open-weight bidirectional encoders, LFM2.5-Encoder-230M and LFM2.5-Encoder-350M, built on the LFM2 hybrid backbone with an 8,192-token context. The 350M model achieves a 17-task…
Liquid AI released LFM 2.5, an 8-billion-parameter agentic model with 1 billion active parameters focused exclusively on tool calling, achieving nearly 200 tokens per second on a GPU. The model excels…
Liquid AI released LFM2.5-Encoder-230M and LFM2.5-Encoder-350M, general-purpose bidirectional encoders that match or beat larger models on GLUE, SuperGLUE, and multilingual tasks while running about 3…
AMD's Advancing AI 2026 summit in San Francisco featured a panel with Chris Lattner, Ramin Hasani, and Hassan Akbari. Lattner, creator of LLVM, called AI 'mid,' arguing that the current AI surface lay…