Agent Breaking Changes_

← Continue

Continue v1.3.35-vscode

2026-03-26 · official notes

Summary

• fix: remove Llama 3.1 405B from Groq provider by @RomneyDa in https://github.com/continuedev/continue/pull/11804
• fix: add keepAlive to YAML completion options schema by @RomneyDa in https://github.com/continuedev/continue/pull/11801
• fix(llama.cpp): include model name in completion request body by @RomneyDa in https://github.com/continuedev/continue/pull/11803

Full notes

## What's Changed
* fix: remove Llama 3.1 405B from Groq provider by @RomneyDa in https://github.com/continuedev/continue/pull/11804
* fix: add keepAlive to YAML completion options schema by @RomneyDa in https://github.com/continuedev/continue/pull/11801
* fix(llama.cpp): include model name in completion request body by @RomneyDa in https://github.com/continuedev/continue/pull/11803
* fix(moonshot): add reasoning_content field support for Kimi models by @RomneyDa in https://github.com/continuedev/continue/pull/11806
* fix: respect contextLength from YAML model config by @RomneyDa in https://github.com/continuedev/continue/pull/11807
* fix(gemini): merge consecutive same-role messages to fix ordering errors by @RomneyDa in https://github.com/continuedev/continue/pull/11805
* chore: prerelease bumps jb 64, vs code 35 by @RomneyDa in https://github.com/continuedev/continue/pull/11834
* fix: lazy-load Ollama /api/show to reduce unnecessary requests by @RomneyDa in https://github.com/continuedev/continue/pull/11836
* fix: don't skip installs by default and sync lock file by @RomneyDa in https://github.com/continuedev/continue/pull/11837
* chore(deps-dev): bump picomatch from 2.3.1 to 2.3.2 in /extensions/cli by @dependabot[bot] in https://github.com/continuedev/continue/pull/11838
* chore(deps): bump yaml from 2.8.1 to 2.8.3 in /packages/hub by @dependabot[bot] in https://github.com/continuedev/continue/pull/11839
* chore(deps-dev): bump picomatch from 2.3.1 to 2.3.2 in /packages/hub by @dependabot[bot] in https://github.com/continuedev/continue/pull/11840
* chore(deps): bump picomatch from 2.3.1 to 2.3.2 by @dependabot[bot] in https://github.com/continuedev/continue/pull/11842
* feat: add `useResponsesApi` option to opt out of Responses API by @RomneyDa in https://github.com/continuedev/continue/pull/11809
* chore(deps-dev): bump picomatch from 2.3.1 to 2.3.2 in /packages/continue-sdk/typescript by @dependabot[bot] in https://github.com/continuedev/continue/pull/11849
* chore(deps-dev): bump picomatch from 4.0.3 to 4.0.4 in /packages/terminal-security by @dependabot[bot] in https://github.com/continuedev/continue/pull/11850
* chore(deps-dev): bump picomatch from 2.3.1 to 2.3.2 in /packages/fetch by @dependabot[bot] in https://github.com/continuedev/continue/pull/11846
* fix: coerce MCP tool args to match schema string types by @RomneyDa in https://github.com/continuedev/continue/pull/11852
* fix: map reasoning-delta to reasoning_content instead of content by @RomneyDa in https://github.com/continuedev/continue/pull/11847
* chore(deps-dev): bump picomatch from 2.3.1 to 2.3.2 in /packages/config-yaml by @dependabot[bot] in https://github.com/continuedev/continue/pull/11844
* chore(deps): bump yaml from 2.8.2 to 2.8.3 in /packages/openai-adapters by @dependabot[bot] in https://github.com/continuedev/continue/pull/11845
* fix: allow multiple context providers of the same type in config.yaml by @RomneyDa in https://github.com/continuedev/continue/pull/11853
* chore(deps): bump picomatch from 2.3.1 to 2.3.2 in /gui by @dependabot[bot] in https://github.com/continuedev/continue/pull/11855
* chore(deps): bump yaml from 2.8.1 to 2.8.3 in /packages/continue-sdk by @dependabot[bot] in https://github.com/continuedev/continue/pull/11859
* feat(openrouter): send HTTP-Referer and X-Title headers to identify app by @RomneyDa in https://github.com/continuedev/continue/pull/11857
* chore(deps-dev): bump picomatch from 2.3.1 to 2.3.2 in /docs by @dependabot[bot] in https://github.com/continuedev/continue/pull/11860
* chore(deps): bump yaml from 2.8.2 to 2.8.3 in /packages/config-yaml by @dependabot[bot] in https://github.com/continuedev/continue/pull/11862
* fix: stop CLI free-trial polling for non-free-trial models by @guoyi-smh in https://github.com/continuedev/continue/pull/11215
* fix: remove inline backtick fences from tool instruction prose by @RomneyDa in https://github.com/continuedev/continue/pull/11869
* fix: handle multiple zip files in JetBrains release artifact step by @RomneyDa in https://github.com/continuedev/continue/pull/11873
* fix: hide thinking indicator for empty thinking content by @RomneyDa in https://github.com/continuedev/continue/pull/11874
* fix(openrouter): Gemini 3 tool support, suffix stripping, thought_signature, autocomplete endpoint by @RomneyDa in https://github.com/continuedev/continue/pull/11856
* fix: fix listener leaks and redundant file read in autocomplete by @RomneyDa in https://github.com/continuedev/continue/pull/11863
* docs: add troubleshooting for Ollama memory errors by @RomneyDa in https://github.com/continuedev/continue/pull/11864
* docs: clarify where secrets can be templated from by @RomneyDa in https://github.com/continuedev/continue/pull/11865
* fix: preserve tool calls when thinking models return no text content by @RomneyDa in https://github.com/continuedev/continue/pull/11866
* fix: allow correcting API key after entering an invalid one for xAI/Gemini by @RomneyDa in https://github.com/continuedev/continue/pull/11854
* fix: show actionable error when Ollama fails to parse tool calls by @RomneyDa in https://github.com/continuedev/continue/pull/11870
* fix: vLLM provider should respect user-configured contextLength and model by @RomneyDa in https://github.com/continuedev/continue/pull/11868

## New Contributors
* @guoyi-smh made their first contribution in https://github.com/continuedev/continue/pull/11215

**Full Changelog**: https://github.com/continuedev/continue/compare/v1.3.34-vscode...v1.3.35-vscode