reallexi/lexi-resume-v6

TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:0.5BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Aug 7, 2026License:otherArchitecture:Transformer0.0K Featherless Exclusive Cold

Reallexi LLC's lexi-resume-v6 is a 495 million parameter language model, fine-tuned from Qwen/Qwen2.5-0.5B-Instruct, specifically designed for resume screening and processing tasks. This specialized model excels at extracting and summarizing information from resumes based on specified job roles. With its merged adapter, it offers efficient deployment for applications requiring targeted resume analysis.

Loading preview...

Overview

reallexi/lexi-resume-v6 is a specialized language model developed by Reallexi LLC, featuring 495 million parameters. It is fine-tuned from the Qwen/Qwen2.5-0.5B-Instruct base model, with its adapter merged into the base weights for streamlined deployment without requiring PEFT at runtime. The model was trained using an slm strategy with Auto LoRA (rank 8, alpha 16) on the AzharAli05/Resume-Screening-Dataset over 10,000 samples across 3 epochs.

Key Capabilities

  • Resume Information Extraction: Optimized to process and extract relevant details from resumes.
  • Role-Based Summarization: Capable of generating summaries tailored to specific job roles provided in the prompt.
  • Efficient Deployment: The merged adapter design simplifies integration into applications.

Use Cases

  • Automated Resume Screening: Ideal for initial filtering and analysis of job applications.
  • Candidate Information Summarization: Quickly generate concise summaries of candidate qualifications for recruiters.
  • HR Technology Integration: Suitable for embedding into HR platforms for enhanced resume processing.

Technical Specifications

  • Parameters: 495M
  • Trained Context Length: 8,192 tokens
  • Memory Footprint: Approximately 944 MB (FP16/BF16), 472 MB (8-bit), 260 MB (4-bit).

Finetunes
1 models