GitHub Adds Cost and Quality Tiers to Copilot Auto Model Selection
Efficiency, balance and intelligence let users express a routing priority, but the selected model still determines usage charges.
Edited by Tyronne Panaino
GitHub added three priorities to Copilot's automatic model selection on September 14: efficiency, balance and intelligence. The control is rolling out in Visual Studio Code, Copilot CLI and the GitHub Copilot app, giving users a way to state how the router should weigh cost, response time and answer quality.
The update matters because automatic selection previously exposed less control over the tradeoff a user wanted for a particular workflow. It still does not let a tier pin one named model or guarantee a particular cost or quality result.
What the three tiers mean
Efficiency prioritizes keeping costs low and is intended for fast, straightforward work. Balance weighs cost, quality and latency together for everyday use. Intelligence prioritizes quality and is positioned for complex tasks.
GitHub says all three tiers draw from the same set of available models. Copilot evaluates each prompt and chooses from that set according to the selected priority. A simple request can therefore route to a smaller model even when the user chose intelligence, because the control adjusts the optimization goal rather than reserving the largest model for every prompt.
Routing preference does not replace billing evidence
Usage remains charged according to the model Copilot actually selects. GitHub also says paid subscribers keep the discount applied to usage billed through automatic selection. Users and administrators still need usage reports or billing records to see which model handled work and what it cost; the tier name alone is not a price quote.
The rollout covers three Copilot surfaces named in the announcement. GitHub does not give a universal completion time, comparative benchmark results for the tiers, or a promise that efficiency will always be cheaper for every workload. Those outcomes depend on prompt-level routing and the models available at the time.
What to watch next
GitHub describes this release as a first step toward more model-selection customization and visibility. The useful next checkpoints are complete availability across the named clients, clearer routing evidence in usage reports and documentation showing how changes to the available model pool affect each tier.
Status
Confirmed. Internal confidence is medium because GitHub's official changelog establishes the tiers, rollout surfaces and billing rule, while comparative cost, latency and quality outcomes remain vendor-framed and unverified here.
Sources
Update note: Last reviewed September 17, 2026. We will revise this article if GitHub changes tier behavior, rollout coverage or billing treatment.
Sources
Drafted with AI assistance from source briefs; reviewed for citation completeness and label accuracy.