tepirale/Ornith-Agents-A1-3.7-35B-A3B-dare_ties_v4

TEXT GENERATIONConcurrent Unit Cost:3Model Size:35.1BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Jul 4, 2026Architecture:Transformer0.0K Featherless Exclusive Cold

The tepirale/Ornith-Agents-A1-3.7-35B-A3B-dare_ties_v4 is a 35.1 billion parameter language model created by tepirale using the DARE TIES merge method. It combines components from InternScience/Agents-A1, Qwen/Qwen3.5-35B-A3B, and deepreinforce-ai/Ornith-1.0-35B, based on Qwen/Qwen3.6-35B-A3B. This model is designed to leverage the strengths of its constituent models, making it suitable for diverse natural language processing tasks.

Loading preview...

Model Overview

The tepirale/Ornith-Agents-A1-3.7-35B-A3B-dare_ties_v4 is a 35.1 billion parameter merged language model. It was constructed by tepirale utilizing the DARE TIES merge method, a technique for combining pre-trained language models.

Merge Details

This model uses Qwen/Qwen3.6-35B-A3B as its base model. The merge process integrated several distinct models to enhance its capabilities:

  • InternScience/Agents-A1: Likely contributes to agentic reasoning or task execution capabilities.
  • Qwen/Qwen3.5-35B-A3B: A foundational Qwen model, providing strong general language understanding.
  • deepreinforce-ai/Ornith-1.0-35B: Suggests contributions related to reinforcement learning or specific domain expertise.

The DARE TIES method, with specific density and weight parameters for each component, was employed to create a synergistic blend of these models. The merging was performed using mergekit and configured for bfloat16 precision.

Potential Use Cases

Given its diverse lineage, this model is potentially well-suited for:

  • Complex reasoning tasks: Benefiting from the 'Agents-A1' component.
  • General-purpose NLP: Leveraging the robust capabilities of the Qwen base models.
  • Applications requiring nuanced understanding: Drawing from the 'Ornith-1.0' component.