DynoTable
Summary
The AWS Console lets you scan a table, stare at raw JSON, and manually piece together relationships that DynamoDB was never designed to surface — DynoTable exists because that workflow breaks the moment your data spans more than one table.
DynoTable is a local-first desktop client that runs SQL with GROUP BY, COUNT DISTINCT, and multi-table JOINs directly against your live DynamoDB data — no ETL, no intermediate server, nothing routed through a third party. An AI agent powered by your own Amazon Bedrock credentials reads your schema, picks Query over Scan where it can, and surfaces proposed writes as a reviewable diff you sign off on before anything commits. Exports of million-plus row result sets run at constant memory. The ceiling appears fast if you need browser-based access, API integration, or a self-hosted deployment — the tool is a desktop app with no API surface exposed.
Bottom line: Pick DynoTable when your engineers need SQL joins and AI-assisted data exploration against DynamoDB without routing credentials through a cloud vendor's servers — reach for something else the moment your workflow requires programmatic access or a browser-based shared interface.
Pricing Plans
Subscription- Price
- $9/month (Individual) or $18/seat/month (Team)
- Free Tier
- Read-only: visual browser, PartiQL console and export only
Free
Read-only access: visual browser, PartiQL console, export
- Visual table browser & PartiQL console
- Multiple AWS profiles & SSO
- Keyboard-first navigation
- Constant-memory export
Individual
Everything in Free plus SQL workbench, editing and AI agent
- SQL workbench with JOINs, GROUP BY, aggregates
- Item editing as JSON with staged commits
- AI agent on your own Bedrock keys
- MCP staging & write tools
Team
Everything in Individual with shared workspace for up to 20 seats
- Centralized seat management & billing
- Shared AI context and table indexes
- Shared data in your own S3 bucket
- Access controlled by your own IAM
Enterprise
Everything in Team plus SSO, SCIM, volume pricing and priority support
- Single sign-on (SAML/OIDC) & SCIM
- Volume pricing & invoicing
- Priority support & security review
- MSA & DPA on request
View full pricing on dynotable.com →
Pricing may have changed since last verified. Check the official site for current plans.
Community Performance Report Card
No community ratings yet. Be the first to rate this tool!
Community Benchmarks Community
Sign in to submit a benchmarkNo community benchmarks yet. Be the first to share a real-world data point.
Pros
Sign in to edit- SQL with real GROUP BY, COUNT DISTINCT, JOINs, and aggregations runs directly against live DynamoDB data, so you stop exporting to a spreadsheet every time a query crosses two tables.
- The AI agent runs on Bedrock credentials you supply — prompts and schema never reach DynoTable's servers — which means you get AI-assisted data exploration without adding a third-party data processor to your compliance review.
- Every write the AI agent drafts surfaces as a reviewable diff you approve before it commits, so a misread prompt cannot silently mutate production data.
- Query plan previews show Scan vs Query, the index chosen, and an RCU estimate before execution, which means you catch table-scan cost surprises before they appear on your AWS bill.
- Constant-memory exports handle result sets the vendor describes as one million or more rows, so large data pulls stop failing mid-export on memory-constrained machines.
Cons
Sign in to edit- There is no API surface — workflows that require programmatic or automated access to DynoTable's query layer have no integration path, and teams building pipelines that need to trigger queries from application code will hit this wall immediately and reach for a different tool.
- The client is desktop-only with no browser interface and no self-hosted option, which means teams whose security policy restricts production credential use to controlled server environments cannot use this tool in that context — at that point, a cloud-based or self-hostable alternative becomes the only viable path.
- MCP-based external agent connections (Claude Code, Cursor, Codex) scope writes to staging and require your approval before commit, which is the right default for safety but adds a manual step that breaks fully automated write pipelines — teams expecting fire-and-forget automation will need to design around it or abandon the MCP integration entirely.
Community Reviews
Sign in to write a reviewNo reviews yet. Be the first to share your experience.
About
- Platforms
- Desktop (Windows, macOS, Linux implied by local-first desktop app)
- API Available
- No
- Self-Hosted
- No
- Last Updated
- 2026-07-28T20:17:01.516Z
Best For
Who it's for
- Engineers working in one or two AWS accounts
- Users who require SQL on DynamoDB without moving data to servers
- Organizations wanting AI assistance on their own Bedrock infrastructure
What it does well
- Querying and analyzing DynamoDB data with SQL and JOINs
- Editing DynamoDB items locally with staged commits
- Running an AI agent for data exploration using private Bedrock models
- Exporting large result sets to JSON/NDJSON/CSV without high memory use
Integrations
Discussion Community
Sign in to commentNo discussion yet. Sign in to start the conversation.
Compare DynoTable
Spotted incorrect or missing data? Join our community of contributors.
Sign Up to ContributeCommunity Notes & Tips Community
Sign in to contributeBe the first to contribute. General notes, observations, gotchas, and tips from people who use this tool day-to-day.
Frequently Asked Questions
- Is DynoTable free?
- DynoTable has a permanent free tier alongside paid upgrades (paid plans from $9/month (Individual) or $18/seat/month (Team)). You can keep using a baseline version indefinitely without paying.
- Is DynoTable open source?
- No — DynoTable is a closed-source tool. Source code is not publicly available.
- What platforms does DynoTable support?
- DynoTable is available on: Desktop (Windows, macOS, Linux implied by local-first desktop app).
Hours Saved & ROI Stories Community
Sign in to contributeBe the first to contribute. Concrete time/cost savings, with context. e.g. "Cut my code review backlog from 4h to 45m per week."
Curated lists that include this category
DynamoDB’s access-pattern model means the AWS Console is fine for looking up a single item and useless for anything that crosses table boundaries. DynoTable is a desktop client that sits on top of your existing AWS credentials — SSO, aws-vault, or DynamoDB Local — and adds a SQL Workbench that runs real SELECT statements with aggregations, multi-table joins, and HAVING clauses against your live data. Every query previews its execution plan, showing Scan vs Query, the index it will use, and an RCU estimate before a single row is read. Results land in the same data grid across all features, so you can sort, filter, open rows, and export without switching contexts.
The differentiating feature is the AI agent, and the architecture matters: the agent runs entirely inside your AWS account via Amazon Bedrock, using credentials you supply. Prompts, schema, and row data never leave your AWS boundary — DynoTable’s servers are not in the loop at all. The agent inspects keys and GSI shapes before running anything, preferring targeted Query operations over full-table Scans. When it drafts a write, it produces a reviewable diff. Nothing commits until you approve it. The vendor describes this as ‘drafts you approve’ — you stay in the loop on every mutation.
Smart Tables adds a visual canvas for drawing relationships between DynamoDB tables and working a joined view live. Join on the primary key or any GSI; add or remove an edge and the grid recomputes. The base tables are never mutated — the joined view is a read-only projection you can export or query. Exports of large result sets — the vendor cites one million or more rows — run at constant memory, which matters when cloud console exports quietly consume enough RAM to crash your session.
The tool is a paid-only desktop application with a free read-only tier that persists after the trial period. There is no API surface — external agents connect through MCP (Claude Code, Cursor, Codex are listed), scoped per profile, with writes still landing in staging for your review before commit. Self-hosting is not an option; if your compliance posture requires running the client on controlled infrastructure rather than an engineer’s machine, that is a hard blocker.