{"slug": "qwen-agentworld", "name": "Qwen-AgentWorld", "category": "ai-models", "category_name": "AI models", "description": "", "recommenders": 3, "commenters": 6, "warners": 0, "score": 3.628, "rising": 0.35, "first": "2026-06-23", "latest": "2026-08-14", "why": "AgentWorld is a pretty recent MoE release from the Qwen folks with quite a bit better performance than 3.6", "why_is_quote": true, "has_verdict": false, "sentiment": {"recommend": 2, "neutral": 3, "compare": 1, "warn": 1}, "spark": [0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0], "listed": true, "loved": [], "caveat": "", "short_quote": "AgentWorld is a pretty recent MoE release from the Qwen folks with quite a bit better performance than 3.6", "kind": "", "icon": false, "family": "qwen", "domain": "qwenlm.ai", "github_repo": null, "official_url": "", "pricing": "", "aliases": ["Qwen-AgentWorld", "Qwen-AgentWorld-35B-A3B", "Qwen AgentWorld"], "verdict": null, "praise": [{"label": "", "summary": "", "quotes": [{"id": 788854608, "item_id": 49303413, "url": "https://news.ycombinator.com/item?id=49303413", "by": "SwellJoe", "date": "2026-08-14", "version": "35B", "stance": "recommend", "quote": "AgentWorld is a pretty recent MoE release from the Qwen folks with quite a bit better performance than 3.6", "excerpt": "AgentWorld is a pretty recent MoE release from the Qwen folks with quite a bit better performance than 3.6, released around the same time as 3.7 Plus/Max, I guess. Not sure why it didn't get more attention, as it is definitely better than 3.6 35B A3B on all dimensions, but especially for agentic use. Still nowhere near the dense models (even 3.6 27B), but clearly an upgrade in the small Qwen MoE line. https://hugging", "context": "", "use_case": "agentic coding model", "switched_from": "", "story": "Qwen3.8-27B"}, {"id": 779564512, "item_id": 48722782, "url": "https://news.ycombinator.com/item?id=48722782", "by": "verdverm", "date": "2026-06-29", "version": null, "stance": "recommend", "quote": "Qwen's new AgentWorld model is good too", "excerpt": "Qwen's new AgentWorld model is good too: https://huggingface.co/Qwen/Qwen-AgentWorld-35B-A3B I'm running the NVFP4 alongside Gemma4 at the same quant on an OEM Spark", "context": "", "use_case": "", "switched_from": "", "story": "Qwen 3.6 27B is the sweet spot for local development"}, {"id": 778669408, "item_id": 48666838, "url": "https://news.ycombinator.com/item?id=48666838", "by": "jedisct1", "date": "2026-06-24", "version": null, "stance": "recommend", "quote": "FOr Mac users: https://huggingface.co/collections/jedisct1/qwen-agentworld", "excerpt": "FOr Mac users: https://huggingface.co/collections/jedisct1/qwen-agentworld", "context": "", "use_case": "", "switched_from": "", "story": "Qwen-AgentWorld: Language World Models for General Agents"}]}], "caveats": [{"label": "", "summary": "", "quotes": [{"id": 778586080, "item_id": 48661630, "url": "https://news.ycombinator.com/item?id=48661630", "by": "khimaros", "date": "2026-06-24", "version": null, "stance": "warn", "quote": "these work but seem to loop in reasoning a lot", "excerpt": "that particular quant is just corrupted. these work but seem to loop in reasoning a lot https://huggingface.co/groxaxo/Qwen-AgentWorld-35B-A3B-GGUF", "context": "", "use_case": "", "switched_from": "", "story": "Qwen-AgentWorld: Language World Models for General Agents"}]}], "who": [], "tags": [], "use_cases": ["agentic coding model", "agentic tasks", "language world model for agents", "agent world model"], "compared": [], "makers": [], "threads": [{"url": "https://news.ycombinator.com/item?id=48654351", "title": "Qwen-AgentWorld: Language World Models for General Agents", "mentions": 6}, {"url": "https://news.ycombinator.com/item?id=49299684", "title": "Qwen3.8-27B", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=48721903", "title": "Qwen 3.6 27B is the sweet spot for local development", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=48709744", "title": "Ornith-1.0: Self-Scaffolding LLMs for Agentic Coding", "mentions": 1}], "sentiment_by_month": {"2025-11": {}, "2025-12": {}, "2026-01": {}, "2026-02": {}, "2026-03": {}, "2026-04": {}, "2026-05": {}, "2026-06": {"neutral": 3, "warn": 1, "recommend": 1, "compare": 1}, "2026-07": {}, "2026-08": {"recommend": 1}, "2026-09": {}, "2026-10": {}}, "versions": [], "version_unnamed": 5, "family_ref": {"slug": "qwen", "name": "Qwen"}, "runs_on": null, "lines": [], "quotes": [{"id": 788854608, "item_id": 49303413, "url": "https://news.ycombinator.com/item?id=49303413", "by": "SwellJoe", "date": "2026-08-14", "version": "35B", "stance": "recommend", "quote": "AgentWorld is a pretty recent MoE release from the Qwen folks with quite a bit better performance than 3.6", "excerpt": "AgentWorld is a pretty recent MoE release from the Qwen folks with quite a bit better performance than 3.6, released around the same time as 3.7 Plus/Max, I guess. Not sure why it didn't get more attention, as it is definitely better than 3.6 35B A3B on all dimensions, but especially for agentic use. Still nowhere near the dense models (even 3.6 27B), but clearly an upgrade in the small Qwen MoE line. https://hugging", "context": "", "use_case": "agentic coding model", "switched_from": "", "story": "Qwen3.8-27B"}, {"id": 779564512, "item_id": 48722782, "url": "https://news.ycombinator.com/item?id=48722782", "by": "verdverm", "date": "2026-06-29", "version": null, "stance": "recommend", "quote": "Qwen's new AgentWorld model is good too", "excerpt": "Qwen's new AgentWorld model is good too: https://huggingface.co/Qwen/Qwen-AgentWorld-35B-A3B I'm running the NVFP4 alongside Gemma4 at the same quant on an OEM Spark", "context": "", "use_case": "", "switched_from": "", "story": "Qwen 3.6 27B is the sweet spot for local development"}, {"id": 779376048, "item_id": 48711003, "url": "https://news.ycombinator.com/item?id=48711003", "by": "SwellJoe", "date": "2026-06-28", "version": null, "stance": "compare", "quote": "did worse than Qwen AgentWorld", "excerpt": "I added this to a benchmark I've been doing of how well agents find security bugs, specifically security bugs originally found by Mythos. It performs poorly with only read/grep/ls tools, but in a follow-up test with a full shell and Python, it doubled its findings (still a poor showing, but it does at least indicate it is doing what it says on the tin: making tools to help it solve problems). It also did worse than Q", "context": "", "use_case": "Agentic tasks", "switched_from": "", "story": "Ornith-1.0: Self-Scaffolding LLMs for Agentic Coding"}, {"id": 778684144, "item_id": 48667759, "url": "https://news.ycombinator.com/item?id=48667759", "by": "khimaros", "date": "2026-06-24", "version": null, "stance": "neutral", "quote": "they have arrived", "excerpt": "they have arrived: https://huggingface.co/unsloth/Qwen-AgentWorld-35B-A3B-GGUF", "context": "", "use_case": "", "switched_from": "", "story": "Qwen-AgentWorld: Language World Models for General Agents"}, {"id": 778669408, "item_id": 48666838, "url": "https://news.ycombinator.com/item?id=48666838", "by": "jedisct1", "date": "2026-06-24", "version": null, "stance": "recommend", "quote": "FOr Mac users: https://huggingface.co/collections/jedisct1/qwen-agentworld", "excerpt": "FOr Mac users: https://huggingface.co/collections/jedisct1/qwen-agentworld", "context": "", "use_case": "", "switched_from": "", "story": "Qwen-AgentWorld: Language World Models for General Agents"}, {"id": 778586080, "item_id": 48661630, "url": "https://news.ycombinator.com/item?id=48661630", "by": "khimaros", "date": "2026-06-24", "version": null, "stance": "warn", "quote": "these work but seem to loop in reasoning a lot", "excerpt": "that particular quant is just corrupted. these work but seem to loop in reasoning a lot https://huggingface.co/groxaxo/Qwen-AgentWorld-35B-A3B-GGUF", "context": "", "use_case": "", "switched_from": "", "story": "Qwen-AgentWorld: Language World Models for General Agents"}, {"id": 778506208, "item_id": 48656638, "url": "https://news.ycombinator.com/item?id=48656638", "by": "kakugawa", "date": "2026-06-24", "version": null, "stance": "neutral", "quote": "So the world model generates the current state (an html document)", "excerpt": "Here's the demo: https://docs.qwenlm.ai/resources/mlu56_demo.html Here's the description of the world model prompt for the web domain: \"A precise GUI state simulator \u2014 given the current screen (as HTML) and a user action, predicts the exact next screen as a complete, self-contained HTML document.\" (You can click the world model prompt box to expand it and see the full prompt.) So the world model generates the current", "context": "", "use_case": "language world model for agents", "switched_from": "", "story": "Qwen-AgentWorld: Language World Models for General Agents"}, {"id": 778502896, "item_id": 48656431, "url": "https://news.ycombinator.com/item?id=48656431", "by": "adrian_b", "date": "2026-06-24", "version": "35B", "stance": "neutral", "quote": "The smaller of the two models is open weights and available on Huggingface", "excerpt": "The smaller of the two models is open weights and available on Huggingface: https://huggingface.co/Qwen/Qwen-AgentWorld-35B-A3B", "context": "", "use_case": "", "switched_from": "", "story": "Qwen-AgentWorld: Language World Models for General Agents"}, {"id": 778483072, "item_id": 48655192, "url": "https://news.ycombinator.com/item?id=48655192", "by": "verdverm", "date": "2026-06-23", "version": null, "stance": "neutral", "quote": "35B model from the qwen-3.5 line", "excerpt": "35B model from the qwen-3.5 line https://github.com/QwenLM/Qwen-AgentWorld https://huggingface.co/Qwen/Qwen-AgentWorld-35B-A3B", "context": "", "use_case": "agent world model", "switched_from": "", "story": "Qwen-AgentWorld: Language World Models for General Agents"}]}