- October 9, 2026 (2026-10-09)
- Changelog We ship fast ⚡
- Blog Thoughts on data
- vs Hyperquery
Compare Deepnote...
- More powerful Deepnote Agent, enhanced project governance, new Runs page & more
Deepnote Agent now works as a single agent from your first instruction to the finished notebook, keeping the same context the whole way through. You'll see clearer progress indicators and the tool calls it makes as it goes, and prompt caching has improved, which can lower AI cred
- End-to-end workflows with Deepnote Agent
Deepnote Agent now works as a single agent from your first instruction to the finished notebook, keeping the same context the whole way through. You'll see clearer progress indicators and the tool calls it makes as it goes, and prompt caching has improved, which can lower AI cred
- Auto AI is now part of Deepnote Agent
Deepnote Agent now works as a single agent from your first instruction to the finished notebook, keeping the same context the whole way through. You'll see clearer progress indicators and the tool calls it makes as it goes, and prompt caching has improved, which can lower AI cred
- Enhanced project governance
Deepnote Agent now works as a single agent from your first instruction to the finished notebook, keeping the same context the whole way through. You'll see clearer progress indicators and the tool calls it makes as it goes, and prompt caching has improved, which can lower AI cred
- Private and shared projects can live in the same folder
Deepnote Agent now works as a single agent from your first instruction to the finished notebook, keeping the same context the whole way through. You'll see clearer progress indicators and the tool calls it makes as it goes, and prompt caching has improved, which can lower AI cred
- Set a default project permission per folder
Deepnote Agent now works as a single agent from your first instruction to the finished notebook, keeping the same context the whole way through. You'll see clearer progress indicators and the tool calls it makes as it goes, and prompt caching has improved, which can lower AI cred
- Latest models, lower prices
Deepnote Agent now works as a single agent from your first instruction to the finished notebook, keeping the same context the whole way through. You'll see clearer progress indicators and the tool calls it makes as it goes, and prompt caching has improved, which can lower AI cred
- API & MCP updates
Deepnote Agent now works as a single agent from your first instruction to the finished notebook, keeping the same context the whole way through. You'll see clearer progress indicators and the tool calls it makes as it goes, and prompt caching has improved, which can lower AI cred
- VS Code extension update
Deepnote Agent now works as a single agent from your first instruction to the finished notebook, keeping the same context the whole way through. You'll see clearer progress indicators and the tool calls it makes as it goes, and prompt caching has improved, which can lower AI cred
- See all your notebook runs in one place
Deepnote Agent now works as a single agent from your first instruction to the finished notebook, keeping the same context the whole way through. You'll see clearer progress indicators and the tool calls it makes as it goes, and prompt caching has improved, which can lower AI cred
- Python 3.14
Deepnote Agent now works as a single agent from your first instruction to the finished notebook, keeping the same context the whole way through. You'll see clearer progress indicators and the tool calls it makes as it goes, and prompt caching has improved, which can lower AI cred