Deepnote
Collaborative data notebook with an AI agent and an open file format
Overview
Deepnote is a hosted data notebook for teams, with real-time co-editing, native SQL blocks, scheduled runs, publishable data apps and the Deepnote Agent. Its .deepnote file format, converter and local toolkit are open source under Apache 2.0, while the cloud service itself remains proprietary.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 14 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- Limited Deepnote AI: 10 completions and 5 agent calls a month
- Unlimited Basic machines (5 GB RAM, 2 vCPU)
- 7-day revision history
- Unlimited viewers and notebooks
- Premium integrations, scheduled notebooks and background execution
- $39 AI credits, $280 CPU and $50 GPU a month
- 30-day revision history
- SSO, directory sync and audit logs
- Unlimited AI and bring your own LLM
- Single-tenancy and private Docker images
- Dedicated success manager
What you can produce with Deepnote
- Notebooks several people edit at the same time, with comments and revision history
- SQL blocks querying warehouses such as Snowflake, BigQuery and Redshift beside Python code
- Scheduled and background notebook runs
- Notebooks published as interactive data apps and dashboards
- Analyses drafted, edited and explained by the Deepnote Agent
- Notebooks converted between .ipynb and the open .deepnote format
- Deepnote notebooks run locally in VS Code, Cursor or Windsurf
Deepnote is Jupyter rebuilt for teams: live co-editing, SQL beside Python, scheduling, publishable apps, and an agent that now carries a task from instruction to finished notebook. Opening the .deepnote format and converter under Apache 2.0 in late 2025 lowers lock-in, though the cloud service where collaboration happens is still proprietary.
The Free plan is usable for students and solo analysts; Team at $39 per editor a month billed yearly ($49 monthly) adds integrations, scheduling and monthly compute and AI credits. Sustained heavy compute costs extra, and teams with data-residency rules should check where execution runs. A strong choice for data teams that do not want to run notebook infrastructure.
What we found
Deepnote is a hosted data notebook built for teams: several people edit the same notebook at once, SQL sits in native blocks beside Python, notebooks can be scheduled or run in the background, and finished work can be published as an app for colleagues who never open a notebook. Deepnote Inc. is a Y Combinator company from the summer 2019 batch, listed in San Francisco, with engineering roots in Prague. In July 2024 it acquired Hyperquery, another analytics startup.
The open-source story needs stating precisely. In November 2025 Deepnote published its notebook format and building blocks on GitHub under the Apache 2.0 licence: the human-readable .deepnote YAML format, the @deepnote/blocks package that defines block types, the @deepnote/convert tool that moves notebooks between .ipynb and .deepnote, and a Python toolkit for running them locally, plus an extension for VS Code, Cursor and Windsurf. The hosted Deepnote Cloud, where collaboration, scheduling and managed compute live, is still proprietary. The README lists a local version of the cloud interface and bring-your-own AI keys as coming, not shipped.
AI is now the centre of the product. The October 2026 changelog merged Auto AI into a single Deepnote Agent that works from first instruction to finished notebook, and moved the model line-up to newer OpenAI and Anthropic versions at lower prices. An August 2026 post introduced an Agent Workspace for turning trusted analyses into reusable skills and apps.
The Free plan allows three editors and five projects on 5 GB machines, with ten AI completions and five agent calls a month. Team costs $39 per editor a month billed yearly or $49 monthly, and includes monthly allowances of $39 in AI credits, $280 of CPU and $50 of GPU time. Enterprise is quoted and adds SSO, audit logs, single-tenancy and bring-your-own LLM. Heavy compute beyond the allowances is billed on top, and for ASEAN teams it is an English-language US service with no regional hosting option published.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
About this listing
This entry was compiled from publicly available data including Deepnote's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Deepnote unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Deepnote directly →
Spotted something out of date? Suggest an update →
Related reading
Science
Eighteen Years of Whale Calls Show Rare Omura's Whales Favour Warm, Food-Rich Waters off A...
Health & Wellness
Fine-Particle Air Pollution Linked to Higher Hospital Admission Risk for Up to 21 Days, St...
Health & Wellness
A Guide for New Mothers Missed Its Main Goal but Nearly Doubled Primary Care Visits
Alternatives to Deepnote
More in Research & Data
