by Abhinavexists · AI Tool · ★ 52
lev is an open System One decision model and the harness that measures it. Give it a state (text, a ticket, an email, or JSON) and a set of typed questions (yes/no, choice, score), and it answers all of them in one forward pass, reading each answer from logits it already computed. It returns calibrated probabilities over exactly the options you supplied and generates no tokens. The model is a LoRA adapter on Qwen3.5-4B; the wire protocol is TypeSafe's , so any TypeSafe client works against it by changing the base URL. <img alt="interfaze" src="https://img.shields.io/badge/Builtby-Inte
| Stars | 52 |
| Forks | 3 |
| Language | Python |
| Category | AI Tool |
| License | Apache-2.0 |
| Quality Score | 61.9112871945127/100 |
| Open Issues | 1 |
| Last Updated | 2026-10-07 |
| Created | 2026-09-20 |
| Platforms | python |
| Est. Tokens | ~18k |
These tools work well together with lev for enhanced workflows:
Explore other popular ai tool tools:
lev is An open System One decision model. It is categorized as a AI Tool with 52 GitHub stars.
lev is primarily written in Python. It covers topics such as jev, modal, qwen.
You can find installation instructions and usage details in the lev GitHub repository at github.com/Abhinavexists/lev. The project has 52 stars and 3 forks, indicating an active community.
lev is released under the Apache-2.0 license, making it free to use and modify according to the license terms.