Deepnote

Collaborative cloud notebooks

Deepnote provides cloud-based notebooks that support SQL, Python, and Markdown blocks. Teams can explore data, store knowledge, and collaborate in real time with commenting and versioning. Notebooks can be scheduled to run hourly, daily, weekly, or monthly, and can be deployed as APIs.

Streamline data exploration and team collaboration in a single workspace.

Turn analyses into interactive apps

Deepnote lets you create and host interactive data apps and dashboards with live data. You can arrange content into custom layouts, hide code blocks, add input blocks and buttons, and instantly visualize dataframes as no-code configurable charts. These apps can be shared with your team via a link.

Deliver polished, interactive reports and dashboards without extra engineering effort.

Build and run autonomous data agents

Deepnote supports building deterministic or non-deterministic agents that can automate insights and actions. Agents can run skills, query data, train models, and deploy them as live endpoints. The Agent Workspace provides a headless environment for agents, with a UI for your team.

Automate repetitive data tasks and enable AI-driven decision making.

100+ integrations via API or MCP

Deepnote connects to your existing data stack with 100+ integrations, including popular warehouses, databases, and lakehouses like Snowflake, BigQuery, Redshift, PostgreSQL, MySQL, Databricks, and more. It also supports dbt metadata, Spark & Snowpark for terabyte-scale data, and drag-and-drop CSV uploads.

Access and analyze data from your entire stack without leaving the workspace.

Enterprise-grade security and compliance

Deepnote is SOC 2 Type II and HIPAA compliant, and GDPR-ready. It offers RBAC, SSO via SAML or OIDC, directory sync, audit logs, static IPs, SSH tunnels, and single-tenant or private cloud deployments. Data is encrypted and you control access to sensitive data.

Maintain control over data access and meet regulatory requirements.

Flexible compute and model deployment

Deepnote provides a range of machines from basic (5 GB RAM, 2 vCPU) to high-memory (128 GB RAM, 16 vCPU) and GPU instances (60 GB RAM, 12 GB K80 GPU). You can train models, serve them directly, and deploy notebooks as APIs. Volume discounts are available for enterprise.

Scale compute resources as needed and put models into production easily.

• Unified data workspaceDeepnote combines notebooks, data apps, dashboards, spreadsheets, pipelines, and models in one workspace, with 100+ integrations to existing data stacks via API or MCP.
• Agent-ready platformDeepnote supports building and running deterministic or non-deterministic agents, with headless components for agents and a UI for teams, plus the ability to launch data agents from tools like Claude Code.
• Flexible pricing tiersFree plan includes up to 3 editors, 5 projects, limited Deepnote AI, and basic machines with 5 GB RAM and 2 vCPU. Team plan at $39/editor/month (yearly) adds unlimited viewers/notebooks, GPT-5.5 and Sonnet 4.6 access, background execution, scheduled notebooks, and monthly AI/CPU/GPU credits.
• Enterprise-grade securityEnterprise plan offers custom contracts, SSO, audit logs, private Docker images, bring your own LLM, single-tenancy, and volume machine discounts, with SOC2 and HIPAA compliance mentioned.

• Pricing detailsThe Team plan price is shown as $39 per editor/month billed yearly, with a 20% discount indicated, but exact monthly billing price is not shown. Enterprise pricing is custom and requires contacting sales.
• AI model accessTeam plan includes GPT-5.5 and Sonnet 4.6 access, but Free plan has limited Deepnote AI (10 completions per month, 5 calls per month for auto AI, generate, edit, explain code). Enterprise offers unlimited AI and bring your own LLM.
• Compute resourcesFree plan includes unlimited basic machines with 5 GB RAM and 2 vCPU. Team plan includes $280 worth of CPU and $50 worth of GPU per month, plus more powerful machines, but specific machine specs are not listed.
• IntegrationsDeepnote claims 100+ integrations through API or MCP, but specific integrations are not enumerated on the pricing page. The main site mentions integrations with data sources, but exact list is not provided in the scraped content.

Build Autonomous Data Agents

Deepnote lets you build and run deterministic or non-deterministic agents directly in your data workspace. For example, you can create a fraud detection agent that summarizes and cancels fraudulent transactions by country, generating SQL blocks to flag risky transactions, training a model, deploying it as a live API, and visualizing results on a chart — all within one project.

Answer Questions with AI

Use Deepnote AI to ask questions about your data in plain language. The platform can run SQL queries, analyze results, and even provide insights like 'Treatment users skipped onboarding 2.4x more often' with driver analysis. You can add follow-up cells to explore further and explain drivers behind metrics such as churn rate.

Generate Dashboards & Data Apps

Turn quick explorations into interactive dashboards and data apps. Deepnote allows you to arrange blocks into custom layouts, hide code, add input blocks and buttons, and publish live apps. Examples include a PLG dashboard tracking signups and activation, or a pipeline-by-segment view shared with your revenue team.

Start Locally, Continue in the Cloud

Deepnote integrates with your local development environment. You can write code in Visual Studio Code or Claude Code, then launch data agents from anywhere. Use the command palette to open a notebook in the cloud, run it as a skill, or schedule it — with kernel state, packages, env vars, and files preserved for a seamless transition.

Define the exact google Colab AI Tools workflow Deepnote should support.

Compare it with closely related AI tools in the same category before committing.

Set review rules for accuracy, privacy, brand voice, compliance, and final approval.

Connect useful outputs to the wider stack instead of leaving them inside the AI tool.