Kilo Code v4.98.2 - v.4.103.0: Teams + Enterprise Updates, Privacy Controls, and Provider Expansions
Hey y’all, Tirumari here with a new week’s worth of updates for Kilo Code! We’ve added a couple of big features to our Teams and Enterprise Plans, added a Zero Data Retention setting, and added several new updates for our providers!
Teams + Enterprise Updates
Two big updates for our Teams and Enterprise plans!
Usage Statistics
First; your Organization’s dashboard now has a Usage Statistics Panel that gives you a quick summary of your usage at a glance, but also an in-depth page that you can access by clicking on “View Detailed Usage.” There you can access Total Cost, Requests, and Tokens, a graph of Cost over time, breakdowns by Provider and Model, and Usage by Day and User! Who doesn’t love granular data?! (Sign up for Teams here!)
Modes Selection
Your organization can now create new modes and send them to your extension!
You can add modes or override existing ones, and you can edit them from your organization dashboard;
Zero Data Retention
I’m sure you, like most folks, are interested in protecting your privacy. Well, you’re in luck, because you can now enable Zero Data Retention under your Provider Routing settings, ensuring no conversation data is stored by Kilo Code and OpenRouter, covering a wide range of models!
Make sure to set it up today if this is something you’ve been looking for.
Revert all Changes
We also got ourselves a bonafide Revert All Changes Button! Now you’ll have a much easier way to undo large tasks, no hassle involved!
Fixes!
Our team added automatic checkpoints before every tool call to prevent data loss
autocomplete now properly honors .kilocodeignore
We also removed the custom autocomplete provider to prevent high costs from unrestricted usage
Model Updates
Supernova 1M context window; a 5x upgrade from 200K! Read more on our blog!
Contributor eliasto set up OVHcloud AI Endpoints as a provider with custom base URL support
Contributor billycao added Synthetic as a provider
Contributor b3nw added dynamic Chutes loading
Performance & UX
Contributor mcowger added chat timestamps for better conversation tracking
We added a setting to hide small costs to declutter cost display
We also added a configurable API timeout for Ollama and LM Studio
Roo Merge
We merged in Roo Code v3.28.8-v3.28.15 this week, and thank you to the team over at Roo! Some highlights;
The cost limit tracking properly resets when clicking “Reset and Continue”
Got some activation improvements to the prompts settings save button
Claude Sonnet 4.5 compatibility updates and model ID fixes across Bedrock, Vertex AI, and Anthropic providers
Improved tool-use adherence and reduced “No tool used” errors
Cleaner prompt output by removing unnecessary tags
Initial ask now properly included in conversation condensing