Local AI recap: 28 new open models in October, million-token agents on one GPU

vramkickedin · reddit · 2026-11-02

A Reddit roundup of local/open AI releases from October 2026, with 28 LLMs plus multimodal entries, all linked:

• Ornith-1.5-122B: million-token agent tasks on a single GPU
• KRAFTON Raon-1-mini-preview: Korean agentic AI on one RTX 5090
• Tencent Hy-Memory-30B-A3B: durable agent memory at 3B active params
• Intern-S2-397B / Sage-A1.5: scientific reasoning and agent orchestration
• deepseek-v4.1-flash-gguf (by antirez): DeepSeek Flash on Apple Silicon
• Indra-OCR: reads PDFs and handwriting at 44 pages/sec on CPU
• MiniMax-M3.1-Flash: multimodal agentic AI with 1M context
• LFM2.5-24B-A2B (LiquidAI): on-device agent reasoning at 2B active
• Also: uncensored GLM-5.3 at 273GB, LingBot-World 2.0 at 16 FPS, ThreadShelf local chat archiving, and more

Related event: Reddit Roundup: 28 New Open-Source AI Models From October, Including Single-GPU Million-Token Agents(2 posts)→

Original post →

More from Models

• OpenAI merges ChatGPT's chat and Work tabs; Codex now shares the same usage pool — TechCrunch AI · 2026-11-02
• Reddit users report Gemini Argon underperforming Gemini 3.1 Pro on instructions and audio — PreparationOdd430 · 2026-11-02
• Dev warning: avoid yearly prepaid LLM plans as labs quietly cut quotas — HankYeomans · 2026-11-02
• KVARK Sparkle 1.0 open-weights: Qwen3.8-27B derivative cuts thinking tokens by up to 67% — dzulajla98 · 2026-11-02
• 129 Cyber Incidents in 10 Days: Open Models Already Causing "Billions per Day" in Damages — Afinetheorem · 2026-11-02
• Rumor: Qwen 4 will be Alibaba's last open-weight model release — Chemical_Hawk_6307 · 2026-11-02

Models channel →