--- format: "aidr-story-markdown/v1" id: "3eac8bc0c60595a0f59a518503d69fc1e170d99ebcd7ed36cadf487e62dc8713" canonical_url: "https://aidr.today/3eac8bc0?lang=en" title: "webAI's 3.66B TwIL-LM3-Pro Beats VibeThinker by 35% in Formal Logic" lang: "en" requested_lang: "en" available_langs: ["en","vi"] translation_fallback: null fallback_fields: [] published_at: "2026-10-01T01:23:02.000Z" category: "Releases" topics: ["reasoning"] source_urls: ["https://marketbrief.now/ai/webais-366b-twil-lm3-pro-beats-vibethinker-by-35percent-in-formal-logic-15af793d","https://huggingnews.com/ai/webais-366b-twil-lm3-pro-beats-vibethinker-by-35percent-in-formal-logic-15af793d"] summary: "Austin-based webAI released its latest reasoning model for local hardware, introducing an open-source tool designed for checking deductions and critiquing proofs. The TwIL-LM3-Pro model features 3.66 billion parameters and leads other small-scale reasoners in formal logic, scoring 35% higher than China's VibeThinker-3B, 24% higher than Qwen3.5-4B, and 47% higher than Liquid AI's LFM2.5-8B-A1B. On the GSM8K evaluation for math word problems, the model scored 94.3%, nearly matching the 97.7% result of the gpt-oss-120b which has 33 times more parameters. The model is distributed as a 2.09GiB Q4 build that runs via llama.cpp on CPUs or local GPUs to keep private data on the device. This release follows 500,000 downloads of the company's first-generation models within a single month. webAI is now developing Meridian, a series of frontier-class on-device models to be available through the webAI application." --- # webAI's 3\.66B TwIL\-LM3\-Pro Beats VibeThinker by 35% in Formal Logic > [Open the canonical story]() **Published:** 2026-10-01T01:23:02.000Z **Category:** Releases **Topics:** reasoning ## Summary Austin\-based webAI released its latest reasoning model for local hardware, introducing an open\-source tool designed for checking deductions and critiquing proofs\. The TwIL\-LM3\-Pro model features 3\.66 billion parameters and leads other small\-scale reasoners in formal logic, scoring 35% higher than China's VibeThinker\-3B, 24% higher than Qwen3\.5\-4B, and 47% higher than Liquid AI's LFM2\.5\-8B\-A1B\. On the GSM8K evaluation for math word problems, the model scored 94\.3%, nearly matching the 97\.7% result of the gpt\-oss\-120b which has 33 times more parameters\. The model is distributed as a 2\.09GiB Q4 build that runs via llama\.cpp on CPUs or local GPUs to keep private data on the device\. This release follows 500,000 downloads of the company's first\-generation models within a single month\. webAI is now developing Meridian, a series of frontier\-class on\-device models to be available through the webAI application\. ## Sources - [Story source]() - [Story source]()