●Change8

Cline: What's New in March 2026

A summary of all updates, new features, and bug fixes released for Cline during March 2026.

6releases9features16fixes7improvements
← February 2026April 2026 →

New Features in March 2026

  • Added Cline Kanban launch modal in the webview, and the CLI now launches Kanban by default with a migration view.(v3.76.0)
  • Added a toggle to disable feature tips within the chat interface.(v3.76.0)
  • Added repeated tool call loop detection to prevent infinite loops that waste tokens.(v3.76.0)
  • Implement dynamic free model detection for the Cline API(v3.74.0)
  • Add feature tips tooltip displayed during the thinking state(v3.74.0)
  • Added W&B Inference by CoreWeave as a new API provider supporting 17 models(v3.73.0)
  • Added Anthropic Opus 4.6 fast mode variants(v3.72.0)
  • Added GPT-5.4 models for ChatGPT subscription users(v3.71.0)
  • Added a Notification hook for attention and completion boundaries(v3.71.0)

Bug Fixes in March 2026

  • Fixed an issue preventing CLI Kanban from spawning correctly on Windows by enabling shell mode for 'npx.cmd'.(v3.76.0)
  • Replaced the generic error message shown when a user is not logged into Cline with a more specific message(v3.74.0)
  • Fixed alignment issue in ClineRulesToggleModal padding to match ServersToggleModal(v3.74.0)
  • Skipped WebP format processing for GLM and Devstral models running via llama.cpp(v3.74.0)
  • Ensured user-configured context window is respected in LiteLLM getModel() calls(v3.74.0)
  • Ensured explicit model IDs outside the static catalog are honored in the W&B provider(v3.74.0)
  • Added missing Fireworks serverless models and their associated pricing information(v3.74.0)
  • Fixed crashes in the Claude Code Provider when encountering rate limit events, empty content arrays, error results, or unknown content types(v3.73.0)
  • Ensured tool handlers (read_file, list_files, list_code_definition_names, search_files) return graceful errors instead of crashing(v3.73.0)
  • Resolved native tool placeholder interpolation issues within prompts(v3.72.0)
  • Fixed Gemini Flash output token limit to 8192 across all providers(v3.72.0)
  • Fixed unit test path normalization issues specifically on Windows(v3.72.0)
  • Resolved flaky hooks tests occurring on Windows environments(v3.72.0)
  • Fixed Bedrock to correctly handle thinking and redacted_thinking blocks during message conversion and streaming(v3.72.0)
  • Prevented crashes when the list_files or list_code_definition_names functions receive an invalid file path(v3.72.0)
  • Fixed unreliable streamable HTTP MCP reconnects after disconnects(v3.71.0)

Improvements in March 2026

  • Improved latency when working in remote workspaces.(v3.75.0)
  • Added file read deduplication cache to prevent unnecessary repeated file reads(v3.74.0)
  • Improved parallel tool calling support for OpenRouter and Cline providers(v3.73.0)
  • Updated Jupyter Notebook GIFs for better clarity(v3.72.0)
  • Markdown image loading now requires explicit user consent before displaying(v3.72.0)
  • Added a new file, .github/copilot-instructions.md, for coding agents(v3.72.0)
  • Reintroduced the feature toggle for Hooks functionality(v3.72.0)

All Releases in March 2026

v3.76.03 features1 fix

This release introduces the Cline Kanban launch modal in the webview and sets Kanban as the default launch experience for the CLI. Key improvements include new loop detection to save tokens and a setting to disable feature tips in the chat.

v3.75.0

This release focuses on performance enhancements, specifically improving latency for remote workspace usage. No major new features were introduced, and example hooks were removed in favor of directing users to the official documentation.

v3.74.02 features6 fixes

This release introduces dynamic free model detection for the Cline API and adds helpful feature tips tooltips during the thinking state. Several bugs were fixed, including improving error messaging for logged-out users and ensuring correct context window handling in LiteLLM. Performance is also improved via a new file read deduplication cache.

v3.73.01 feature2 fixes

This release introduces W&B Inference by CoreWeave as a new supported API provider with 17 models. Key fixes address stability issues in the Claude Code Provider and ensure tool handlers return graceful errors instead of crashing. Parallel tool calling support has also been improved for OpenRouter and Cline.

v3.72.01 feature6 fixes

This release introduces support for the new Anthropic Opus 4.6 fast mode variants, expanding model options for users. Several critical bugs were fixed, including issues with native tool placeholder interpolation and stability problems on Windows. Additionally, user experience improvements include requiring consent for Markdown image loading.

v3.71.02 features1 fix

This release introduces the highly anticipated GPT-5.4 models for ChatGPT subscribers, significantly expanding available capabilities. Developers will benefit from the new Notification hook for better boundary management. Additionally, stability has been improved with more reliable handling of HTTP MCP reconnects.