GitHub Copilot CLI Can Discover Local Ollama Models

On October 7, 2026, GitHub announced that Copilot CLI can discover compatible models from a running local Ollama instance through the /model command, starting with CLI version 1.0.94-0.

Users can review a model's provider and endpoint before adding it to a session. Ollama and the model must already be installed; discovery does not download either one.

GETTING STARTED: Install Ollama, obtain a compatible model and start its local service. Then open the model selector in a supported Copilot CLI version and inspect the detected model and endpoint. If discovery fails, check the Ollama process, CLI version and model compatibility.

LOCAL INFERENCE IS NOT OFFLINE OPERATION: Local inference means the model computation runs on the user's machine. Offline operation concerns the network behavior of the wider application. Authentication, updates, telemetry or connected tools may still use network services.

PRIVACY CHECKS: Organizations handling sensitive source code should review which files enter prompts, where the CLI connects and where logs are stored. Tests with non-sensitive sample data and observation of network traffic can help validate the actual configuration.

HARDWARE TRADE-OFFS: Model size, quantization and context length affect memory needs and latency. A larger model is not automatically the best choice for every task. Compare correctness, response time and resource usage across representative coding workloads.

LOCAL VERSUS CLOUD: Local models can reduce reliance on external inference services but require local compute and model maintenance. Cloud services may reduce device requirements while introducing contractual and data-handling considerations. Match the deployment approach to the task and sensitivity of the code.

AVOIDING MISCONCEPTIONS: Appearing in the model picker does not guarantee equivalent quality across all Copilot tasks. Nor does local inference automatically disable network communication. Verify the documented settings and observed behavior separately.

TECHNICAL CONTEXT: The announced approach needs to be understood in its specific technical and operational context. A useful evaluation begins by identifying the exact task, the information available to the system and the expected outcome.

IMPLEMENTATION CONSIDERATIONS: The practical value depends on how the system is integrated with existing processes and controls. Teams should identify which actions are permitted, how failures are detected and who can review consequential results.

EVALUATION AND LIMITS: The stated capabilities and figures should be evaluated under their reported conditions. Independent tests and representative real-world tasks help establish whether the approach is suitable beyond a demonstration.

WHAT TO WATCH: The long-term value depends on integration with existing work, cost, reliability and the ability to verify results. Organizations should track real deployments and repeat evaluations as products change, rather than rely solely on initial demonstrations.

Selecting a local model does not automatically enable offline mode or disable GitHub telemetry. GitHub documents a separate COPILOT_OFFLINE=true setting.

For sensitive development work, teams should verify where prompts and code context are sent. Running a local model and operating entirely offline are not the same thing.