Coding Tools
Copilot Cloud Agent Adds Auto Model Selection
GitHub Changeloghigh confidence
Why it matters
- →Auto model selection optimizes model performance based on real-time conditions.
- →Users benefit from cost savings and no weekly rate limits.
- →Enhances the efficiency of AI-driven coding tools for developers.
GitHub has introduced an auto model selection feature for its Copilot cloud agent. This new capability allows the system to automatically choose the best model based on current system health and performance metrics. Users selecting the auto option will receive a 10% discount on the model multiplier and will not face weekly rate limits. This development aims to enhance the efficiency and accessibility of AI-driven coding assistance for developers.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
How we got here
- 1CopilotKit raises $27M for app-native AI agents
TechCrunch AI · May 5, 2026 · Related
What happened next
- 2GitHub Expands Copilot Cloud Agent Model Options
GitHub Changelog · May 18, 2026 · Same story
- 3GitHub Copilot Enhances Model Selection in VS Code
GitHub Changelog · May 20, 2026 · Same story
- 4GitHub Copilot App Now Supports Any Model Provider
Matt Wolfe · June 10, 2026 · Same story
Follow this story
Open the full story →GitHub Introduces Copilot Agent Tasks REST API
3 developments
- May 13 · GitHub Changelog
GitHub Introduces Copilot Agent Tasks REST API - May 14 · GitHub Changelog
GitHub Copilot Enhances JetBrains IDEs with CLI Agent↳ GitHub Copilot integrates its CLI agent into JetBrains IDEs with a unified sessions view and an 'Ask question' tool. - May 14 · GitHub Changelog
Copilot Cloud Agent Adds Auto Model Selection (This article)↳ Copilot Cloud Agent adds auto model selection with 10% discount and no weekly rate limits
More in Coding Tools
Coding Toolscoding
Claude Code v2.1.316: Subagents and Gateway Budgets
Anthropic is tightening control over Claude Code's autonomous behavior with new environment variables that let you disable 'ultrathink' or cap subagent effort levels. For teams, the introduction of pooled group budgets on the gateway allows shared monthly spending limits across members, a practical fix for enterprise cost management. The release also patches critical bugs where shell commands would hang or crash on exit, stabilizing the developer experience significantly.
Claude Code Releases
Coding Toolscoding
llama.cpp b12022 adds ROCm 10 and Snapdragon support
This release quietly cements llama.cpp as the universal inference runtime by finally bringing ROCm 10.0 to Linux and Windows alongside CUDA 13.4, effectively closing the hardware gap for AMD users who were previously left behind. The inclusion of native Snapdragon support on Linux marks a significant step toward running local LLMs on ARM-based mobile chips without emulation overhead. While KleidiAI on Apple Silicon is disabled in this build, the broader expansion to diverse accelerators ensures that developers aren't locked into NVIDIA ecosystems anymore.
llama.cpp Releases
Coding Toolscoding
llama.cpp b12023 fixes shared prefix bug
This release addresses a specific inference artifact where decision prefixes were incorrectly bleeding into output embeddings. By capping shared prefixes at the first output position, llama.cpp ensures that option tokens remain distinct for parent and child tasks. This is a targeted fix for multi-step reasoning or voting workflows rather than a performance boost. It stabilizes complex generation patterns without changing the underlying model architecture.
llama.cpp Releases