A simpler, more powerful experience
We’re making a set of focused improvements to streamline the product and deliver a faster, more intuitive experience. In some cases, this means evolving how certain capabilities work so they’re easier to use, more reliable, and more powerful in practice.
Improved Chat
Chat is being uplifted to become smarter. Rather than try to get everything right in one shot, it can now do thinking and perform more sophisticated context lookup than before.
What this means for your users
For end-users, nothing other than a noticeably improved chat experience. For self-hosted customers, GPT-OSS, Gemma, and Qwen 3 will no longer be strong enough.
When will this happen
Our May release (automatic for EMT, v6.2 for self-hosted)
Changes to the Test experience
The test experience is moving to /test, where it becomes a more flexible, dedicated environment. This gives it the space to be more performant and easier to work with, while keeping the same core capability—just significantly improved.
- More flexible and focused environment
- Better performance and usability
- Cleaner separation from the main workflow
- Greater range of testing frameworks, methodologies, and approaches
Same capability, significantly improved experience.
What this means for your users
Users will stop going to the Testing-Tab in the UI to create new test-plans and instead use the improved Chat functionality to type /test
When will this happen
Our June release (automatic for EMT, v6.3 for self-hosted)
Inline Actions
We’re also simplifying the interface by removing inline actions. As workflows have evolved, the core chat based experience has become the center of how people interact with the product. Streamlining here allows us to reduce unnecessary complexity and focus on making the primary workflow faster, more consistent, and more dependable. And above all, remaining in line with current preferred development trends.
- Most usage has shifted to chat-based workflows
- This allows us to focus on making core interactions faster and more reliable
- Reduces UI complexity and improves consistency
Net result: fewer distractions, better core experience.
What this means for your users
Users still get the functionality of inline actions, but will use chat to access them via the respective commands (e.g. /explain).
When will this happen
Our June release (automatic for EMT, v6.3 for self-hosted)
Code Awareness
Finally, we’re introducing a new approach to code awareness in chat. This is a lighter-weight, more responsive system that removes the need for Docker and eliminates long indexing times. The result is a chat experience that’s immediately useful and more accurate, with a deeper understanding of your code from the start.
- No Docker setup required
- No long indexing times
- Improved accuracy and relevance
- Lighter-weight, faster startup
Chat becomes immediately useful, without waiting or configuration.
What this means for your users
Nothing except their IDE will feel snappier.
When will this happen
Our June release (automatic for EMT, v6.3 for self-hosted)
Summary
These changes are part of a broader effort to reduce friction, improve performance, and focus on what matters most. The result is a product that feels simpler, faster, and more capable across the board.
