appvoid/cloud-6

TEXT GENERATIONConcurrent Unit Cost:1Model Size:0.35BQuant:BF16Context Size:32kPublished:Jul 6, 2026Architecture:Transformer Featherless Exclusive Cold

appvoid/cloud-6 is a 0.35 billion parameter language model created by appvoid through a DARE TIES merge of MihaiPopa-1/LFM2.5-350M-heretic, squ11z1/claude-oss-350m, and appvoid/palmer-005-core. This model combines conversational instruction-following capabilities with improved generalization and synthetic behavior, making it suitable for diverse small-scale language generation tasks. It features a 32768-token context length, offering extended processing capacity for its size.

Loading preview...

Overview

appvoid/cloud-6 is a 0.35 billion parameter language model developed by appvoid, created using the DARE TIES merge method via mergekit. This model integrates capabilities from several base models to achieve a balanced performance profile.

Merge Details

The model was constructed using MihaiPopa-1/LFM2.5-350M-heretic as the base. It incorporates two distinct models, each contributing specific characteristics:

  • squ11z1/claude-oss-350m: Contributes to assistant tone and conversational instruction-following abilities.
  • appvoid/palmer-005-core: Enhances generalization and synthetic behavior.

Both merged models were weighted equally (0.50) with a density of 0.50 during the DARE TIES process, which included int8_mask: true and normalize: false parameters. The model supports a substantial context length of 32768 tokens.

Key Capabilities

  • Conversational Instruction-Following: Designed to respond with an assistant-like tone and follow instructions effectively.
  • Enhanced Generalization: Benefits from the inclusion of appvoid/palmer-005-core for broader applicability.
  • Synthetic Behavior: Capable of generating diverse and synthetic text outputs.

Use Cases

This model is well-suited for applications requiring a compact yet capable language model that can handle both conversational interactions and general text generation tasks, particularly where a balance between instruction adherence and creative output is desired.