AyoubChLin/LFM2.5-1.2B-Instruct-Saudi-Dialect-1

TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:1.2BQuant:BF16Context Size:32kPublished:Feb 24, 2026Architecture:Transformer Featherless Exclusive Cold

AyoubChLin/LFM2.5-1.2B-Instruct-Saudi-Dialect-1 is a 1.17 billion parameter instruction-tuned language model developed by AyoubChLin, based on Liquid AI's LFM2.5-1.2B-Instruct architecture. This model features a 32,768 token context length and is specifically fine-tuned for Saudi dialect conversational generation. It excels at producing authentic Saudi dialect responses, making it ideal for chatbots, customer support, and conversational agents requiring regional Arabic fluency.

Loading preview...

Saudi Dialect LFM2.5 - Instruction-Tuned Arabic Dialect Model

This model, developed by AyoubChLin, is a specialized fine-tuned version of Liquid AI's LFM2.5-1.2B-Instruct, designed for Saudi dialect conversational generation. The base LFM2.5 architecture is a hybrid state-space + attention model with approximately 1.17 billion parameters and a 32,768 token context length, known for its fast on-device inference and low memory usage. This variant significantly improves fluency, authenticity, and instruction following for regional Arabic use cases by specializing in Saudi dialect conversational patterns.

Key Capabilities

  • Authentic Saudi Dialect Generation: Produces natural and fluent responses in Saudi Arabic.
  • Fast and Efficient Inference: Designed for quick processing and low memory footprint, suitable for edge devices and CPUs.
  • Strong Conversational Coherence: Maintains consistent and relevant dialogue flow.
  • Instruction Following: Responds effectively to given instructions in a conversational context.

Good For

  • Saudi Dialect Chatbots: Creating conversational agents that interact naturally in Saudi Arabic.
  • Customer Support Assistants: Developing support systems tailored for Saudi Arabian users.
  • Arabic NLP Research: Investigating dialect-specific language processing.
  • Dialect-aware RAG Pipelines: Enhancing retrieval-augmented generation systems with regional dialect understanding.
  • Dialogue Generation Systems: Building systems that generate authentic Saudi dialect conversations.

Finetunes
1 models