# Up to 3.2x Faster Inference with LFM2.5-DSpark

DevFeed: [Up to 3.2x Faster Inference with LFM2.5-DSpark](<https://devfeed.tech/articles/up-to-3-2x-faster-inference-with-lfm2-5-dspark-7017.md>)

Original publisher: [Read original article](<https://huggingface.co/blog/LiquidAI/lfm25-dspark>)

Author: Xx; Leonie Monigatti; Fernando Fernandes Neto; Tarek Dakhran; Nathan Ranchin

Published: 2026-08-20T16:52:57Z

Content type: article

Language: en

Sources: [Hugging Face - Blog](<https://devfeed.tech/sources/hugging-face-blog.md>)

Topics: [Inference](<https://devfeed.tech/topics/inference.md>), [Large Language Model](<https://devfeed.tech/topics/llm.md>), [Latency](<https://devfeed.tech/topics/latency.md>), [Streaming](<https://devfeed.tech/topics/streaming.md>), [sglang](<https://devfeed.tech/topics/sglang.md>), [GPU](<https://devfeed.tech/topics/gpu.md>), [Code](<https://devfeed.tech/topics/code.md>)

Tags: [agentic](<https://devfeed.tech/tags/agentic.md>), [benchmark](<https://devfeed.tech/tags/benchmark.md>), [code](<https://devfeed.tech/tags/code.md>), [gpu](<https://devfeed.tech/tags/gpu.md>), [inference](<https://devfeed.tech/tags/inference.md>), [latency](<https://devfeed.tech/tags/latency.md>), [llama](<https://devfeed.tech/tags/llama.md>), [llama-cpp](<https://devfeed.tech/tags/llama-cpp.md>), [llm](<https://devfeed.tech/tags/llm.md>), [memory](<https://devfeed.tech/tags/memory.md>), [on-device](<https://devfeed.tech/tags/on-device.md>), [streaming](<https://devfeed.tech/tags/streaming.md>)

## AI overview

The article presents LFM2.5-DSpark, a speculative decoding approach that accelerates LLM inference by using a lightweight draft model and verifier. It reports up to 3.18x GPU throughput improvement and up to 2.87x on-device improvement, along with reduced function-calling latency and support for llama.cpp and SGLang.

## Source excerpt

- Faster inference: up to 3.18 throughput improvement on a GPU and up to 2.87x on-device. - Toward on-device agentic inference: cuts function-calling latency by 57% on average for LFM2.5-2.6B - Day-one support for llama.cpp and SGLang: LFM-compatible DSpark integration is open-sourced upstream The decode phase in LLM inference is traditionally memory-bound. Most latency comes from streaming weights from DRAM into SRAM, not from intense computation.