AI Tool Updates - June 6, 2026

Week of: June 6, 2026

---

Overview

This week saw a flurry of releases across AI developer tools, with Ollama shipping multiple patch versions adding new model support and platform fixes, Claude Code introducing fallback model configuration, and GitHub deprecating older GPT models in Copilot while expanding enterprise plugin management.

Stories

1. Ollama v0.30.6 Adds Gemma 4 QAT Models and MLX Improvements

Source: Ollama Link: https://github.com/ollama/ollama/releases/tag/v0.30.6

Ollama released v0.30.6, introducing Gemma 4 QAT (Quantization-Aware Training) weights across multiple model sizes (2B, 4B, 12B, 26B, and 31B) to reduce memory requirements and maximize on-device performance. The update also integrates ollama launch omp with Oh My Pi, an AI coding agent with IDE integration, and improves MLX embedding layers using NVFP4 global scale for better quantization on Apple Silicon.

This release continues Ollama's rapid iteration cycle, with three patch versions (v0.30.4, v0.30.5, v0.30.6) shipping in quick succession. The QAT model support is particularly significant for developers running local models on consumer hardware, as it directly reduces the memory footprint needed to run Gemma 4.

Impact Analysis: Developers running local LLMs should update to v0.30.6 to access memory-optimized Gemma 4 QAT models, especially on Apple Silicon where MLX quantization improvements also apply.

2. Claude Code v2.1.166 Adds Fallback Model Support and Security Hardening

Source: Claude Code Link: https://github.com/anthropics/claude-code/releases/tag/v2.1.166

Claude Code v2.1.166 introduces a fallbackModel setting that allows configuring up to three fallback models tried in order when the primary model is overloaded or unavailable, with --fallback-model now applying to interactive sessions. The release also adds glob pattern support in deny rule tool-name positions, hardens cross-session messaging by stripping user authority from relayed messages, and blocks auto-mode on relayed permission requests.

The fallback model configuration is a practical reliability improvement for CI/CD pipelines and automated workflows where model availability is critical. The security hardening around cross-session messaging addresses a potential privilege escalation vector.

Impact Analysis: Teams using Claude Code in automated workflows should configure fallback models to maintain uptime, and review deny rules for the new glob pattern support.

3. GitHub Deprecates GPT-5.2 and GPT-5.2-Codex in Copilot

Source: GitHub Changelog Link: https://github.blog/changelog/2026-06-05-gpt-5-2-and-gpt-5-2-codex-deprecated

GitHub deprecated GPT-5.2 and GPT-5.2-Codex models across most Copilot experiences, including Copilot Chat, inline edits, ask and agent modes, and code completions, effective June 5, 2026. The deprecation applies to all Copilot surfaces using these model versions.

This signals a model refresh cycle in GitHub Copilot, likely pushing users toward newer model versions. Developers relying on specific behavior from GPT-5.2 should test their workflows against the replacement models.

Impact Analysis: Copilot users should verify their code completion and chat workflows still function as expected and update any pinned model references.

4. Enterprise-Managed Plugins in VS Code Now in Public Preview

Source: GitHub Changelog Link: https://github.blog/changelog/2026-06-05-enterprise-managed-plugins-in-vs-code-in-public-preview

GitHub expanded its enterprise-managed plugins feature from Copilot CLI to VS Code, now in public preview. Enterprise administrators can configure and distribute plugins to GitHub Copilot CLI users and VS Code users across their organization.

This is a significant step for enterprise adoption of AI-assisted development, giving IT teams centralized control over which AI plugins developers can use. The VS Code integration broadens the reach beyond CLI-only workflows.

Impact Analysis: Enterprise teams should evaluate the public preview to establish plugin governance policies before general availability.

5. CodeQL 2.25.6 Adds Swift 6.3.2 Support and Improved C# Coverage

Source: GitHub Changelog Link: https://github.blog/changelog/2026-06-05-codeql-2-25-6-adds-swift-6-3-2-support-and-improves-c-coverage

CodeQL 2.25.6 adds support for Swift 6.3.2 and completes C# coverage improvements. The static analysis engine powers GitHub code scanning for security issue detection.

Swift developers on the latest language version can now benefit from CodeQL's security analysis, while improved C# coverage strengthens .NET security scanning.

Impact Analysis: Teams using GitHub code scanning should update CodeQL to 2.25.6 for Swift 6.3.2 compatibility and better C# vulnerability detection.

6. n8n 2.25.5 Fixes AI Builder Sandbox Persistence

Source: n8n Link: https://github.com/n8n-io/n8n/releases/tag/n8n%402.25.5

n8n 2.25.5 fixes critical issues with Instance AI builder sandboxes, keeping them thread-scoped and non-ephemeral, and resuming sandboxes that were stopped or deleted while idle. The release also includes telemetry fixes for experiment attribution.

These fixes address reliability issues in n8n's AI workflow builder, ensuring sandboxed AI agent executions persist correctly across sessions.

Impact Analysis: n8n users running AI agent workflows should upgrade to 2.25.5 to prevent sandbox state loss during idle periods.

7. LangChain Core 1.4.1 and Perplexity 1.3.2 Released

Source: LangChain Link: https://github.com/langchain-ai/langchain/releases/tag/langchain-core%3D%3D1.4.1

LangChain released core v1.4.1 with updates to model profile docstrings, uuid-utils bumped to 0.16.0, and serialization improvements for message construction. The langchain-perplexity integration also updated to v1.3.2 with ToolMessage and AIMessage.tool_calls support.

These incremental updates improve type safety and serialization for LangChain applications, particularly around tool-calling patterns.

Impact Analysis: LangChain users should update to core 1.4.1 for improved serialization and the latest dependency versions.

Source Links

• Ollama - v0.30.6
• Claude Code - v2.1.166
• GitHub Changelog - GPT-5.2 and GPT-5.2-Codex deprecated
• GitHub Changelog - Enterprise-managed plugins in VS Code in public preview
• GitHub Changelog - CodeQL 2.25.6 adds Swift 6.3.2 support and improves C# coverage
• n8n - n8n@2.25.5
• LangChain - langchain-core==1.4.1