Tralalabs/Nebulos-Distill-Qwen3-0.6B

TEXT GENERATIONConcurrency Cost:1Model Size:0.8BQuant:BF16Ctx Length:32kTool Calling:SupportedPublished:Jan 5, 2026Architecture:Transformer0.0K Cold

Tralalabs/Nebulos-Distill-Qwen3-0.6B is a compact 0.6 billion parameter causal language model, distilled from the Qwen 3 architecture by Erik22TY. Fine-tuned for efficient step-by-step logic and reasoning, it excels at tasks like math word problems and logic puzzles. This model is optimized for high-speed local inference and minimal VRAM requirements, making it suitable for mobile or low-spec desktop environments.

Loading preview...

Nebulos-Distill-Qwen3-0.6B: A Compact Reasoning Model

Nebulos-Distill-Qwen3-0.6B is a 0.6 billion parameter causal language model developed by Erik22TY, distilled from the Qwen 3 architecture. This model was fine-tuned using LoRA on consumer-grade hardware (NVIDIA GTX 1050 with 3GB VRAM) in approximately 1 hour and 15 minutes, demonstrating efficient training on limited resources. It is designed for high-speed local inference and minimal VRAM consumption.

Key Capabilities

  • Efficient Reasoning: Specifically fine-tuned to perform step-by-step logic and maintain logical consistency.
  • Low Resource Footprint: Operates with minimal VRAM, making it suitable for local deployment on mobile devices or low-spec desktop environments.
  • Compact Size: At 0.6 billion parameters, it offers a balance between performance and resource efficiency.
  • Apache 2.0 License: Available for broad use under a permissive open-source license.

Good For

  • Reasoning-heavy tasks: Excels in applications requiring logical problem-solving, such as math word problems and logic puzzles.
  • Concise text generation: Capable of generating focused and brief textual outputs.
  • Local deployment: Ideal for scenarios where inference needs to happen on-device or in environments with limited computational resources.

Limitations

Due to its small size, the model is not recommended for long-form creative writing or highly complex professional advice, as it may be prone to hallucination in such contexts.