appvoid/cloud-3

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

appvoid/cloud-3 is a 0.35 billion parameter language model created by appvoid, merged using the DARE TIES method. It integrates capabilities from OrionLLM/Terminus-LFM2.5-350m, squ11z1/claude-oss-350m, and mkurman/LiquidAI-LFM2.5-350M-SYNTH. This model is designed to combine terminal/agentic behavior, conversational instruction-following, and generalization, making it suitable for diverse small-scale AI applications.

Loading preview...

appvoid/cloud-3: A Merged 0.35B Parameter Model

appvoid/cloud-3 is a compact 0.35 billion parameter language model developed by appvoid. It was created using the DARE TIES merge method, which combines the strengths of several specialized base models into a single, efficient unit. This approach allows the model to inherit diverse capabilities without significantly increasing its parameter count.

Key Capabilities

This model integrates functionalities from three distinct sources:

  • Terminal / Agentic / Tool Behavior: Derived from OrionLLM/Terminus-LFM2.5-350m, providing capabilities for interacting with tools and executing agentic tasks.
  • Conversational Instruction-Following: Enhanced by squ11z1/claude-oss-350m, contributing to a more natural assistant tone and improved conversational abilities.
  • Generalization / Synthetic Behavior: Benefiting from mkurman/LiquidAI-LFM2.5-350M-SYNTH, which aids in broader understanding and synthetic data generation.

Use Cases

Given its merged architecture, appvoid/cloud-3 is well-suited for applications requiring a blend of:

  • Small-scale agentic systems: Where the model needs to perform specific actions or interact with external tools.
  • Conversational AI: For chatbots or virtual assistants that require coherent instruction-following and a helpful tone.
  • General text generation: For tasks benefiting from a model with diverse foundational knowledge and synthetic reasoning capabilities, particularly in resource-constrained environments due to its small size.