Skip to content

Qwythos 9B Claude Mythos 5 1M

empero-ai/Qwythos-9B-Claude-Mythos-5-1M
Open Source · chat · open-weights · seen 14m ago
GA Alert me on changes
Context
1M
Max output
Input $/1M
Output $/1M
Modalities
text
Released
19 Jun 2026
AI summary
● machine-written

Qwythos 9B: 9B parameter reasoning model with 1M context window

Qwythos-9B-Claude-Mythos-5-1M is a 9 billion parameter model fine-tuned from Qwen3.5-9B, trained on over 500 million tokens of Claude Mythos and Fable traces. It features a 1,048,576-token context window via YaRN rope-scaling and native function calling, designed for complex reasoning tasks in technical domains including cybersecurity, biomedical analysis, and quantitative reasoning.

What's new
  • 1M token context window via YaRN rope-scaling for whole-codebase reasoning
  • Native function calling compatible with OpenAI/Qwen3.5 spec
  • +34 MMLU and +30 GSM8K-strict performance gains over base Qwen3.5-9B
  • Self-correction with tool integration for factually accurate responses
  • Hybrid Gated-DeltaNet attention stack keeps memory sub-quadratic below 256k tokens
Best for
Complex multi-step reasoning and problem-solvingLong-context applications including codebase analysis and multi-document researchAgentic workflows with tool integration and retrieval-augmentationTechnical domain expertise in cybersecurity, biomedical, and quantitative fields
Sources

Source: https://huggingface.co/empero-ai/Qwythos-9B-Claude-Mythos-5-1M