Cline v4.1.26 adds cloud task toggle, refreshes model catalog and changes defaults

The release adds a Local | Cloud toggle in Settings > API Configuration when cloud sessions are on, letting cloud tasks use a Cline model saved separately from local models. Starting or resuming a cloud sandbox now shows a progress card listing the repo, branch and steps, which collapses once the agent starts. Cloud sessions also hide the Plan/Act toggle and auto-approve bar rather than showing them disabled.

The model catalog is refreshed, adding the CommandCode, Reflection, Liquid Inference and Munito AI providers and dropping CrofAI, iFlow and Llama. ClinePass now defaults to Kimi K3.1 Pro, and default models change for most other providers. Model requests now retry at most 3 times instead of 5, and Kimi K3 gets max reasoning for Extra High and Max.

Fixes cover GPT-6 models on OpenAI and Azure OpenAI-compatible endpoints, which now receive max_completion_tokens, and Gemini API keys no longer sending native Google Search. Several task-handling issues around Plan/Act switching, cancelled questions and auto-compaction are also addressed, along with PowerShell escaping guidance for the run_commands tool.

If you use a provider without pinning a model in Cline, the default model you get may now be different, so pin the model explicitly if your agent depends on a specific one.

Agents using the gemini provider no longer get native Google Search injected, and models that Cline dropped from its catalog, such as Llama, CrofAI and iFlow, are no longer offered.