Ishowbackup/PINQWEN-3.5-9B-1M-BF16

VISIONConcurrent Unit Cost:1Model Size:9BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Aug 19, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

PINQWEN-3.5-9B-1M is a 9 billion parameter reasoning model developed by Blackfrost AI, based on the Qwen 3.5 architecture. It features an exceptional 1,000,000-token context window, enabling it to process extensive inputs like entire codebases or books. This multimodal model integrates a full vision encoder, allowing it to handle both text and image inputs, and is specifically tuned for software engineering, technical problem-solving, and transparent chain-of-thought reasoning.

Loading preview...

PINQWEN-3.5-9B-1M: A Compact Reasoning Powerhouse

PINQWEN-3.5-9B-1M, developed by Blackfrost AI, is a 9 billion parameter model built on the Qwen 3.5 architecture. It stands out with its 1,000,000-token context window, a feature that allows it to process vast amounts of information, such as entire codebases or extensive documents, efficiently. The model incorporates a full vision encoder, making it capable of understanding and reasoning with both text and image inputs.

Key Capabilities

  • Ultra-Long Context: Processes up to 1 million tokens, designed for speed and efficiency with a YaRN-extended window and gated-linear-attention hybrid backbone.
  • Transparent Reasoning: Utilizes a native <think>…</think> chain-of-thought mechanism, distilled from multiple frontier models, providing inspectable and clear reasoning steps.
  • Multimodal: Supports both text and image inputs within a single 9B model.
  • Uncensored: Engineered to provide direct answers without the reflexive refusals common in over-aligned models, offering users full control over its applications.
  • Optimized for Technical Tasks: Specifically tuned for software engineering, step-by-step problem-solving, and precise technical explanations.

Good For

  • General reasoning tasks.
  • Coding assistance and technical Q&A.
  • Understanding ultra-long documents and entire codebases.
  • Multimodal applications involving both images and text.