GitHub: Copilot for VS Code Gets Terminal Access and Bring-Your-Own API Keys
GitHub Copilot for Visual Studio Code received semantic search across the entire codebase, agent access to open terminals, and the ability to plug in your own API keys for Anthropic, OpenAI and other providers during the April release cycle (versions 1.116–1.119).
This article was generated using artificial intelligence from primary sources.
GitHub published a cumulative changelog for Copilot in Visual Studio Code on Wednesday, covering the April development cycle from version 1.116 to 1.119. The biggest leap came on the agents side: they now have direct access to open terminals, full codebase search, and — for business users — can run on their own external model provider keys.
What do agents now see in the workspace?
Agents see significantly more context than before. Semantic indexing now works across all workspaces and enables search by meaning rather than exact string matching. A new githubTextSearch tool adds grep-style query capability across GitHub repositories and organizations to agent sessions. Agents can also read from and write to foreground terminals, including live REPL sessions and interactive scripts, which effectively closes the loop between code suggestions and their execution.
What changes in the chat interface?
Code changes are now shown as inline diffs directly in the chat thread, so users no longer need to open separate diff views to review suggestions. An experimental /chronicle command adds local chat history search. Additionally, users can share open browser tabs with the agent on demand, giving it a live web page as context.
BYO model and cost optimization
The biggest structural change is BYO model. Copilot Business and Enterprise plan users can connect their own API keys from OpenRouter, Microsoft Foundry, Google, Anthropic, OpenAI and others, and local execution via Ollama and Foundry Local is also supported. In parallel, GitHub introduced smarter prompt caching and deferred tool loading to reduce token consumption during agentic tasks.
Frequently Asked Questions
- What does BYO model mean in Copilot?
- Bring-Your-Own model lets Copilot Business and Enterprise plan users connect their own API keys from OpenRouter, Microsoft Foundry, Google, Anthropic, OpenAI and others, instead of being limited to GitHub's own quotas.
- What kind of terminal access do agents get?
- Agents can now read from and write to existing foreground terminals, including active REPL sessions and interactive scripts, extending the conversational workflow beyond the code editor.
- How does Copilot reduce token costs?
- The April release introduces smarter prompt caching, deferred tool loading and new agent tools that together reduce token consumption during longer agentic tasks.
Sources
Related news
Anthropic: Managed Agents gain multiagent sessions, Outcomes, webhooks and vault refresh in public beta
GitHub: validation of agentic behavior via dominator analysis from compiler theory achieves 100% accuracy vs 82% agent self-assessment
vLLM: Mooncake distributed KV cache store integration delivers 3.8× higher throughput and 46× lower P50 TTFT for multi-turn agentic workloads