Skip to main content

​
A high-performance model, optimized for real-world enterprise workloads

Alfred-ft5 Alfred-ft5 is designed for teams that want the very best performance LightOn can deliver today. It is built on top of a modern 32-billion-parameter vision–language architecture, fine-tuned by LightOn to excel in production RAG and agentic workflows. Beyond raw benchmark scores, Alfred-ft5 focuses on robustness, reliability, and tool-centric behavior: the model is trained and evaluated in realistic enterprise scenarios involving long contexts, multiple tools, and complex task decomposition. Where Alfred-sv5 prioritizes European origin and sovereignty, Alfred-ft5 targets organizations whose primary concerns are quality and capability.

​
Technical elements

In line with LightOn’s philosophy, the LLM is one component of a broader RAG and agentic stack. The real value lies in orchestration and retrieval. Still, for teams that need technical anchors, Alfred-ft5 offers:
  • a context window of up to 32,768 tokens enabling rich, multi-document prompts while keeping latency under control,
  • a 32B-parameter, vision–language architecture capable of understanding both text and images for advanced multimodal workflows,
  • strong instruction-following behavior tuned for enterprise documentation, procedures, and compliance scenarios,
  • optimized reasoning and planning for multi-step tasks, including breaking down complex problems into actionable steps.
LightOn’s toolset ensures the context passed to the LLM is as high-quality as possible, with state-of-the-art parsing models and information retrieval stack, complementing these strengths and making sure Alfred-ft5 shines in RAG-centric and agentic use cases.

​
Best-in-class tool use and agentic behaviors

One of Alfred-ft5’s main differentiators is its tool-calling performance, particularly in agentic setups powered by LightOn: indeed, it offers improved tool selection and argument construction, reducing failed or ambiguous calls, as well as better multi-tool coordination, where the model must call several tools in sequence or in parallel. The model is fine-tuned specifically to support and enhance LightOn’s tool ecosystem. This includes RAG pipelines, function-calling interfaces, and higher-level agents, so that Alfred-ft5 behaves as a reliable, predictable decision-making core instead of a generic chat assistant.

​
Frontier vs Sovereign: two complementary product lines

LightOn structures its offering around two complementary model families:
  • Alfred-ft5: for organizations that prioritize top-tier performance, reasoning, and tool use, independent of model origin.
  • Alfred-sv5: for teams that must meet European-origin or similar sovereignty requirements, sometimes at the expense of using the single most capable global model.
Both families benefit from the same LightOn R&D expertise and evaluation pipelines. This separation lets organizations choose a model line that aligns with their risk, compliance, and performance priorities, while keeping the rest of their RAG and agentic stack unchanged.