August 26th, 2026 Model
Introducing List 3.0 Ultra Coder Brain
Our strongest coding model yet
List 3.0 Ultra Coder Brain is our most capable model for software engineering: repository-scale context, precise edits across files, and tool use tuned for real developer workflows.
It writes, tests, and refactors alongside you — from a single function to a full-stack feature.
Repository-scale reasoning
List 3.0 keeps the whole repository in mind: imports, call graphs, and conventions flow into every suggestion, so generated code fits the codebase instead of fighting it.
Agentic coding loops
Paired with the List Cloud agent runtime, it plans multi-step tasks, runs tests, reads failures, and iterates until green — with every step visible in the trace.
Built for teams
Review-ready diffs, conventional commit messages, and documentation updates come standard, so the model's output lands cleanly in your pull request flow.