Microsoft's Intelligent Terminal 0.2 is its biggest release yet, introducing local model support via BYOM (bring your own model) with Ollama or any OpenAI-compatible server. New slash commands include /agent for running different AI agents per tab, /move for repositioning the Agent Pane, and /sessions for searching conversation history. OpenCode joins GitHub Copilot and Codex as a first-class built-in agent. Agents can now run inside WSL distributions, matching the environment where work actually lives. The Agent Pane gains mouse navigation, prompt history (up to 50 per tab), token usage/cost display, image paste via Alt+V, and animated tool-call states. Autofix now handles PowerShell 5.1 runtime errors, parser errors, Bash prompt cycles, and command-not-found cases. Reliability fixes address tab dragging, mid-response tab closing, and multiline paste. Terminal session restore is teased as the next major feature.
Table of contents
Local model support is here! Copy linkNew slash commands Copy linkOpenCode is now supported Copy linkBring your agent into WSL Copy linkA richer, easier-to-read Agent Pane Copy linkPaste images into the Agent Pane Copy linkSmarter Autofix Copy linkReliability and improvements Copy linkCommunity Copy linkWhat’s next? Copy linkQuestions this post answers
How do I use a local model like Ollama with Microsoft Intelligent Terminal?
In Intelligent Terminal 0.2, open the agent's settings, go to the BYOM (bring your own model) providers section, click Add New, enter a Base URL and Model ID pointing at your local server, save, then select the new provider from the Model dropdown. This works with GitHub Copilot and OpenCode agents and supports any server that speaks the OpenAI API, including Ollama. Developers running local LLMs in their terminal workflows track tool updates like this on daily.dev.
Can I run different AI agents in separate tabs in Intelligent Terminal?
Yes, Intelligent Terminal 0.2 introduces the /agent slash command, which lets each tab run a distinct agent independently. You can have Copilot in one tab, Claude in another, and Codex in a third. Conversations are isolated per tab, and you can pick from installed, policy-allowed agents via a picker or by typing the agent name inline. Teams evaluating which AI coding agent fits which workflow find the comparisons they need on daily.dev.
What does Intelligent Terminal 0.2 show for token usage and cost?
Intelligent Terminal 0.2 adds a token usage and cost display to the agent status bar, visible as you work so you can monitor consumption without leaving the terminal. The display can be toggled on or off in settings. Developers keeping cloud AI costs in check while building terminal workflows stay informed on daily.dev.