jaymanaryan/CleanPool-v2-w30

TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:0.8BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Sep 2, 2026Architecture:Transformer Featherless Exclusive Cold

The jaymanaryan/CleanPool-v2-w30 is a 0.8 billion parameter language model merge, built upon the Qwen3-0.6B base using the DARE TIES method. This model integrates several specialized Qwen3-0.6B variants, including those focused on 'Treatment' and 'Diagnose' tasks, alongside 'dreamwriter' and 'Cerium' models. It is designed to combine diverse capabilities from its constituent models, offering a versatile small-scale language solution with a 32768 token context length.

Loading preview...

Model Overview

The jaymanaryan/CleanPool-v2-w30 is a 0.8 billion parameter language model created by jaymanaryan through a merge of several pre-trained models. It utilizes the DARE TIES merge method, with Qwen/Qwen3-0.6B serving as the foundational base model.

Key Characteristics

This model is a composite of multiple specialized Qwen3-0.6B variants, aiming to consolidate their respective strengths. The merged components include:

  • mrfakename/dreamwriter-0.6b-beta
  • prithivMLmods/Cerium-Qwen3-R1-Dev
  • suayptalha/Qwen3-0.6B-Treatment
  • suayptalha/Qwen3-0.6B-Diagnose
  • prithivMLmods/Nenque-MoT-0.6B-Elite14

Each of these models contributed to the final merge with specific density and weight parameters, as configured in the DARE TIES process. The merge configuration also specified int8_mask: true and dtype: bfloat16.

Potential Use Cases

Given its merged nature from models with names suggesting specific functionalities (e.g., 'Treatment', 'Diagnose', 'dreamwriter'), CleanPool-v2-w30 is likely intended for applications requiring a blend of capabilities. Its small parameter count (0.8B) and substantial context length (32768 tokens) make it suitable for efficient deployment in scenarios where resource constraints are a factor, while still handling moderately long inputs.