{"version":1,"type":"story","url":"https://digestai.news/story/pistis-introduces-27b-and-9b-parameter-multimodal-models-via-new-train","json":"https://digestai.news/story/pistis-introduces-27b-and-9b-parameter-multimodal-models-via-new-train.json","markdown":"https://digestai.news/story/pistis-introduces-27b-and-9b-parameter-multimodal-models-via-new-train.md","slug":"pistis-introduces-27b-and-9b-parameter-multimodal-models-via-new-train","headline":"Pistis introduces 27B- and 9B-parameter multimodal models via new training framework","summary":"Researchers released the Pistis model family on arXiv, featuring two multimodal large language models built on Qwen3.6 and Qwen3.5. The 27B-parameter model uses Qwen3.6, while the 9B-parameter model is based on Qwen3.5. Both leverage a novel post-training framework called Interleaved Distillation and Reinforcement Learning (IDRL), which alternates between on-policy distillation and reinforcement learning for better knowledge transfer and optimization stability.\n\nThe framework produces two specialized variants per scale: Pistis-Thinking for deep multimodal reasoning and Pistis-Agentic for long-horizon planning, iterative reasoning, and tool use. Pistis-Agentic excels in multimodal search. The team also introduced Pistis-Auto-Harnessing (PAH), a system-level method that optimizes inference harnesses without altering model parameters or increasing interaction costs. Experiments show PAH improves performance without parameter updates.","keyPoints":["Pistis model family includes 27B-parameter (Qwen3.6-based) and 9B-parameter (Qwen3.5-based) multimodal LLMs","IDRL framework alternates distillation and reinforcement learning for stronger performance and stability","Pistis-Agentic variant excels in multimodal search and supports long-horizon agentic tasks"],"whyItMatters":"The Pistis framework and models advance multimodal AI capabilities, particularly in reasoning and agentic behavior, with potential applications in search and complex decision-making systems.","category":{"slug":"models","name":"Generative AI & Models","url":"https://digestai.news/category/models"},"entities":{"companies":[],"models":["Pistis","Qwen3.6","Qwen3.5","Pistis-Thinking","Pistis-Agentic"],"people":[]},"firstPublishedAt":"2026-09-25T04:00:00Z","updatedAt":"2026-09-25T04:00:00Z","sourceCount":1,"hasPrimarySource":true,"sources":[{"outlet":"arXiv cs.AI","title":"Pistis Technical Report","url":"https://arxiv.org/abs/2609.28554","publishedAt":"2026-09-25T04:00:00Z","type":"primary","primary":true,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Pistis introduces 27B- and 9B-parameter multimodal models via new training framework\", 25 September 2026, https://digestai.news/story/pistis-introduces-27b-and-9b-parameter-multimodal-models-via-new-train","publisher":"Digest AI","title":"Pistis introduces 27B- and 9B-parameter multimodal models via new training framework","datePublished":"2026-09-25T04:00:00Z","url":"https://digestai.news/story/pistis-introduces-27b-and-9b-parameter-multimodal-models-via-new-train"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}