{"slug": "liquid-ai-launches-sub-20ms-latency-stack", "title": "Liquid AI launches sub-20ms latency stack", "summary": "Liquid AI announced Liquid RIPL (Real-time Inference Pipeline), a low-latency inference engine that returns complete generated model responses end-to-end in under 50 ms while sustaining continuous peak-load traffic. Liquid AI reported that LFM2.5-1.2B on Liquid RIPL achieves 14.4–28.4× lower p95 latency than Qwen3.5-0.8B on vLLM across two real-time workloads, and LFM2.5-2.6B achieves 5.6–7.7× lower p95 latency than Qwen3.5-2B. The engine targets enterprise real-time workloads such as search ranking, product recommendations, fraud decisions and payment authorizations that have millisecond latency budgets.", "body_md": "[News](https://www.liquid.ai/news)\n\n# RIPL: Real-time Intelligence When Every Millisecond Counts\n\nToday we're announcing **Liquid RIPL (Real-time Inference Pipeline)**, a **low-latency inference engine**  built to put Liquid Foundation Models (LFMs) directly in the path of a live request. It returns complete generated model responses end-to-end in **under 50 ms**, and it keeps doing so while traffic arrives continuously at peak load.\n\nMost of today's serving infrastructure is built for a different job: general purpose models, large batches, and long running agentic workflows where a few seconds per request is acceptable. Enterprise real-time workloads have far smaller latency budgets. Search rankings, product recommendations, fraud decisions, payment authorizations, and other real-time workloads need an answer in milliseconds. Even small delays add friction to the user experience, like a slower checkout that hurts conversion. That's why these systems have traditionally used fast, but less accurate, methods like gradient-boosted trees or small embedding models.\n\n**Liquid’s RIPL** changes that trade off. It’s built for real-time constraints, delivering faster, more responsive generative AI experiences that improve how customers search, discover, transact, and interact. \n\n## Results across real-time workloads\n\nIn Figure 1, we highlight a few different use cases for small LLMs in real time enterprise workflows, enabled by low-latency LLM inference.\n\nAcross the two workloads in Figure 2, LFM2.5-1.2B on **Liquid RIPL** has **14.4–28.4× lower p95 latency** than Qwen3.5-0.8B on vLLM. LFM2.5-2.6B has **5.6–7.7× lower p95 latency** than Qwen3.5-2B.", "url": "https://wpnews.pro/news/liquid-ai-launches-sub-20ms-latency-stack", "canonical_source": "https://www.liquid.ai/blog/low-latency", "published_at": "2026-10-22 22:11:16+00:00", "updated_at": "2026-10-22 22:41:12.463292+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-infrastructure", "ai-products"], "entities": ["Liquid AI", "Liquid RIPL", "Liquid Foundation Models", "LFM2.5-1.2B", "LFM2.5-2.6B", "Qwen3.5-0.8B", "Qwen3.5-2B", "vLLM"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/liquid-ai-launches-sub-20ms-latency-stack", "markdown": "https://wpnews.pro/news/liquid-ai-launches-sub-20ms-latency-stack.md", "text": "https://wpnews.pro/news/liquid-ai-launches-sub-20ms-latency-stack.txt", "jsonld": "https://wpnews.pro/news/liquid-ai-launches-sub-20ms-latency-stack.jsonld"}}