ModelsSelected73°

webAI releases TwIL-LM3-Pro model

Why we picked it

This model demonstrates efficient parameter utilization for specialized logical reasoning, making advanced AI capabilities accessible on consumer hardware. Developers working on formal verification systems or educational tools can leverage this lightweight model for local deployment without cloud dependencies.

webAI has released TwIL-LM3-Pro, a 3.66B parameter model specialized in formal logic.

Body

The model matches Qwen3-8B performance on formal-logic benchmarks with less than half the parameters.It outperforms VibeThinker-3B on all six formal logic tasks and requires only 2.

09 GiB of storage.

Original · kimmonismus

webAI has released a 3.66B model for formal logic that can run locally on a laptop.

In webAI’s tests, TwIL-LM3-Pro is roughly level with Qwen3-8B on the headline formal-logic composite, with less than half the parameters. It also leads the public VibeThinker-3B checkpoint on all six reported formal tasks.

The team specialized IBM’s Granite 4.2 through post-training to check whether conclusions follow from premises, translate statements into logic and critique formal proofs.

The Q4 GGUF weight file is just 2.09 GiB and can run locally via llama.cpp. You can keep inference on your own machine.

Weights and model card (non-commercial license): https://t.co/GQA7TQlE40

@thewebai @Davidstout

Industry Coordinates & Source Verification

This report is categorized under AITOP's "Models" desk. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include TwIL-LM3-Pro, webAI, Qwen3-8B, local deployment.

Source attribution: curated from kimmonismus (@kimmonismus).