Ollama v0.32.0 adds Qwen3.5 support and an agent UI

Ollama — a popular tool for running large language models (LLMs) locally — shipped v0.32.0 on July 11, 2026, focused on new model support and an agent-oriented experience.

Key points

• Adds a parser and renderer for the Qwen3.5/Next model family, so Ollama can run and format their output correctly.
• Introduces an agent UI for the cmd command, moving agent-style workflows into the CLI.
• Adds a warning before launching older "agent" models, helping avoid accidentally using outdated ones.
• v0.32.0 follows the v0.32.0-rc0 candidate (released July 10), rolling these changes into the release.

This is a feature update aimed at self-hosted LLM users; see the official GitHub releases page for the full changelog and upgrade notes.