Cline v3.86.0 Adds Claude Opus 4.8 and Moonshot Kimi K2.6 Support
Cline v3.86.0 adds provider support for Claude Opus 4.8, including 1M-context variants where available, and Moonshot Kimi K2.6 models. The update also fixes prompt-cache support display for Qwen 3.7 Max.
Cline v3.86.0 Adds Claude Opus 4.8 and Moonshot Kimi K2.6 Support
Cline, the AI coding assistant VS Code extension, released v3.86.0 on May 28, 2025, adding support for two new model providers.
New Model Support
The release adds provider integration for Claude Opus 4.8, including 1M-context variants where available. This follows Anthropic's recent release of the Claude 4 series, bringing extended context capabilities to the Cline coding workflow.
Cline also now supports Moonshot AI's Kimi K2.6 model, expanding its range of available LLM backends beyond the existing OpenAI, Anthropic, and other providers.
Technical Fixes
The update resolves a display issue where prompt-cache support for Qwen 3.7 Max was not properly shown in the Cline provider interface. The prompt caching feature allows for more efficient API usage by reusing previously processed context.
Additionally, v3.86.0 fixes startup permissions in the VS Code nightly publish workflow, addressing a deployment pipeline issue.
Project Restructuring
The development team moved the VS Code extension project into a new apps/vscode directory structure, indicating a shift toward a monorepo organization. This architectural change typically enables better code sharing and development tooling across multiple application targets.
What This Means
Cline's rapid integration of Claude Opus 4.8 — within days of its release — demonstrates the competitive pressure among AI coding assistants to support the latest models. The 1M-context support is particularly relevant for developers working with large codebases who need the assistant to understand extensive file hierarchies and dependencies. The addition of Moonshot Kimi K2.6 continues Cline's strategy of provider diversity, giving users options beyond the dominant OpenAI and Anthropic ecosystems, which may offer cost or performance advantages for specific use cases.
Related Articles
Cline v4.1.24 defaults to Claude Sonnet 5.5 and adds Claude Haiku 5.5 to its model catalog
Cline v4.1.24 makes Claude Sonnet 5.5 the default model for the Cline and OpenRouter providers when no model is selected. The release also adds Claude Haiku 5.5 and three new providers, and fixes reasoning-off 400 errors on several models plus an auto-compaction token-count bug.
Google builds Antigravity into Gemini app for macOS, limited to AI Ultra subscribers
Google has integrated Antigravity, its agent-first development platform, directly into the Gemini app for macOS as part of the Gemini 4 Argon rollout. Access requires an AI Ultra subscription, while the new model itself requires AI Pro or above. The update also renames and reorganizes several app features, including Gemini Spark, Skills and Connectors.
Google DeepMind: Gemini 4 Argon scores 87% on ERQA embodied reasoning, targets full robotics stack
Google DeepMind says Gemini 4 Argon scores 87% on the Embodied Reasoning (ERQA) benchmark using pure reasoning, ahead of every other frontier model. A new robotics post demonstrates robot control, Real2Sim reconstruction, physical understanding, and self-improvement workflows. Pricing, context window, and availability were not disclosed.
Gemini 4 Argon expands to Google AI Pro subscribers; API priced at $2/$10 per 1M tokens
Google is extending Gemini 4 Argon from AI Ultra to AI Pro subscribers in the Gemini app and AI Mode. The Gemini API costs $2 per 1M input tokens and $10 per 1M output tokens until November 30, 2026, then doubles to $4 and $20.
Comments
Loading...