Liquid AI Releases LFM2.5: A Compact AI Mannequin Household For Actual On Gadget Brokers

Liquid AI has launched LFM2.5, a brand new technology of small basis fashions constructed on the LFM2 structure and centered at on machine and edge deployments. The mannequin household consists of LFM2.5-1.2B-Base and LFM2.5-1.2B-Instruct and extends to Japanese, imaginative and prescient language, and audio language variants. It’s launched as open weights on Hugging Face and uncovered by the LEAP platform.

Structure and coaching recipe

LFM2.5 retains the hybrid LFM2 structure that was designed for quick and reminiscence environment friendly inference on CPUs and NPUs and scales the info and put up coaching pipeline. Pretraining for the 1.2 billion parameter spine is prolonged from 10T to 28T tokens. The instruct variant then receives supervised tremendous tuning, choice alignment, and enormous scale multi stage reinforcement studying centered on instruction following, device use, math, and data reasoning.

Textual content mannequin efficiency at one billion scale

LFM2.5-1.2B-Instruct is the primary normal function textual content mannequin. Liquid AI staff stories benchmark outcomes on GPQA, MMLU Professional, IFEval, IFBench, and several other perform calling and coding suites. The mannequin reaches 38.89 on GPQA and 44.35 on MMLU Professional. Competing 1B class open fashions corresponding to Llama-3.2-1B Instruct and Gemma-3-1B IT rating considerably decrease on these metrics.

On IFEval and IFBench, which goal multi step instruction following and performance calling high quality, LFM2.5-1.2B-Instruct stories 86.23 and 47.33. These values are forward of the opposite 1B class baselines within the above Liquid AI desk.

Japanese optimized variant

LFM2.5-1.2B-JP is a Japanese optimized textual content mannequin derived from the identical spine. It targets duties corresponding to JMMLU, M-IFEval in Japanese, and GSM8K in Japanese. This checkpoint improves over the overall instruct mannequin on Japanese duties and competes with or surpasses different small multilingual fashions like Qwen3-1.7B, Llama 3.2-1B Instruct, and Gemma 3-1B IT on these localized benchmarks.

Imaginative and prescient language mannequin for multimodal edge workloads

LFM2.5-VL-1.6B is the up to date imaginative and prescient language mannequin within the sequence. It makes use of LFM2.5-1.2B-Base because the language spine and provides a imaginative and prescient tower for picture understanding. The mannequin is tuned on a variety of visible reasoning and OCR benchmarks, together with MMStar, MM IFEval, BLINK, InfoVQA, OCRBench v2, RealWorldQA, MMMU, and multilingual MMBench. LFM2.5-VL-1.6B improves over the earlier LFM2-VL-1.6B on most metrics and is meant for actual world duties corresponding to doc understanding, consumer interface studying, and multi picture reasoning below edge constraints.

Audio language mannequin with native speech technology

LFM2.5-Audio-1.5B is a local audio language mannequin that helps each textual content and audio inputs and outputs. It’s introduced as an Audio to Audio mannequin and makes use of an audio detokenizer that’s described as eight occasions sooner than the earlier Mimi primarily based detokenizer on the similar precision on constrained {hardware}.

The mannequin helps two fundamental technology modes. Interleaved technology is designed for actual time speech to speech conversational brokers the place latency dominates. Sequential technology is aimed toward duties corresponding to automated speech recognition and textual content to speech and permits switching the generated modality with out reinitializing the mannequin. The audio stack is educated with quantization conscious coaching at low precision, which retains metrics corresponding to STOI and UTMOS near the total precision baseline whereas enabling deployment on units with restricted compute.

Key Takeaways

• LFM2.5 is a 1.2B scale hybrid mannequin household constructed on the LFM2 machine optimized structure, with Base, Instruct, Japanese, Imaginative and prescient Language, and Audio Language variants, all launched as open weights on Hugging Face and LEAP.
• Pretraining for LFM2.5 extends from 10T to 28T tokens and the Instruct mannequin provides supervised tremendous tuning, choice alignment, and enormous scale multi stage reinforcement studying, which pushes instruction following and power use high quality past different 1B class baselines.
• LFM2.5-1.2B-Instruct delivers sturdy textual content benchmark efficiency on the 1B scale, reaching 38.89 on GPQA and 44.35 on MMLU Professional and main peer fashions corresponding to Llama 3.2 1B Instruct, Gemma 3 1B IT, and Granite 4.0 1B on IFEval and IFBench.
• The household consists of specialised multimodal and regional variants, with LFM2.5-1.2B-JP attaining state-of-the-art outcomes for Japanese benchmarks at its scale and LFM2.5-VL-1.6B and LFM2.5-Audio-1.5B protecting imaginative and prescient language and native audio language workloads for edge brokers.

Take a look at the Technical details and Model weights. Additionally, be happy to observe us on Twitter and don’t neglect to hitch our 100k+ ML SubReddit and Subscribe to our Newsletter. Wait! are you on telegram? now you can join us on telegram as well.

Take a look at our newest launch of ai2025.dev, a 2025-focused analytics platform that turns mannequin launches, benchmarks, and ecosystem exercise right into a structured dataset you’ll be able to filter, examine, and export

The put up Liquid AI Releases LFM2.5: A Compact AI Model Family For Real On Device Agents appeared first on MarkTechPost.