DynoTable vs DocKit
DocKit is one of the few desktop clients that open both Amazon OpenSearch and DynamoDB. It is free and open source, and it also covers Elasticsearch, MongoDB and EasySearch. DynoTable covers two engines, OpenSearch and DynamoDB, and goes deeper on both: it signs OpenSearch requests with your AWS profile, stages document edits as a diff you review, and runs SQL that joins an index to a DynamoDB table.
Pick DocKit when you need one free client across many datastores. Pick DynoTable when your data lives in AWS and you want OpenSearch and DynamoDB in the same workspace, joined in one query.
Which one fits your work
| Your workflow | Choose |
|---|---|
| One free client for Elasticsearch, MongoDB, OpenSearch and DynamoDB | DocKit |
| Write REST requests against any of 120+ cluster API endpoints | DocKit |
| Configure DynamoDB streams or table encryption | DocKit |
| A self-hosted cluster that requires a password or an API key | DocKit |
| Open an IAM-only OpenSearch domain or a Serverless collection | DynoTable |
| Review a document edit as a diff and revert it later | DynoTable |
JOIN an index to a DynamoDB table, or GROUP BY over the joined rows | DynoTable |
| Let an AI agent change data only through changes you approve | DynoTable |
How each client signs in to Amazon OpenSearch
DocKit's OpenSearch connection takes a username and password or an API key.
We read its source at the current commit (8782620, 28 October 2026): the
connection sends a Basic or ApiKey header and nothing else. AWS request
signing lives only in its DynamoDB client, and AWS profile and SSO sign-in is
an Ultimate feature.
That works for a domain with fine-grained access control and an internal user.
It does not cover a domain whose access policy only admits IAM principals, or a
Serverless collection, whose data access policies grant IAM roles and users.
AWS's client examples for collections all sign each request with Signature
Version 4 for the aoss service
(AWS docs).
DynoTable signs every OpenSearch request with the credentials of your AWS profile, on every plan, and finds your domains and collections with them. You add no connection by hand. A self-hosted cluster takes an endpoint URL per profile and connects with no credentials at all, which is where DocKit is the better tool: a self-hosted cluster that requires a password or an API key connects in DocKit and not in DynoTable.
Searching an index
DocKit gives OpenSearch a Monaco editor for REST requests, with JSON5, inline comments, field completion from the live mapping and a catalog of more than 120 API endpoints. Results show as a table, a tree or JSON. It works like the Dev Tools console in OpenSearch Dashboards, in a desktop window.
DynoTable opens an index as a grid tab. A Lucene search box completes field names from the mapping and values from the index, filter pills compile to query DSL, and a time range appears when the index has a date field. Raw mode takes a JSON query DSL body or a Lucene line, linted against the mapping, with saved queries and per-index history. Column sorts run on the server, and a free-text search highlights the matches in each cell.
DynoTable has no generic REST console. A request that isn't a search, such as a reindex or a snapshot call, still belongs in a REST tool.
Editing documents
DocKit edits a document from its result table: Edit opens the JSON in a
dialog, and saving sends PUT /<index>/_doc/<id> with refresh=true. The
write is immediate, and the request carries no if_seq_no, so a change someone
made after you loaded the row is overwritten without a warning.
DynoTable stages the edit. Pending changes shows a per-field diff, and you commit all of it or part of it. On a domain, the update carries the version of the document you read, so a concurrent change stops the commit with a conflict you can rebase. Serverless collections have no version check, and DynoTable says so with a Last-write-wins badge on the tab. Every commit is journaled locally and can be reverted from History. Bulk edit and bulk delete run over a selection or every document a query matches, staged the same way.
SQL across OpenSearch and DynamoDB
DocKit keeps each data source in its own editor. Its grammar catalog includes
POST /_plugins/_sql, so you can send SQL to the OpenSearch SQL plugin, with
the plugin's limits. OpenSearch documents its joins as best-effort with no
aggregations on the joined result, and when we ran a JOIN … GROUP BY through
the plugin on OpenSearch 2.19 it returned one row per document with null
counts and no error. The OpenSearch Dashboards comparison
shows the request and the response.
DynoTable's SQL Workbench runs the join in the app. One statement can join an index to a DynamoDB table, or to another index on a different domain:
SELECT c.country, COUNT(*) AS orders, SUM(o.amount) AS revenue
FROM dynamodb.orders o
JOIN opensearch.customers c ON o.customerId = c.customerId
GROUP BY c.countryAn index joins on a keyword, number or date field. On the DynamoDB side, a join onto a table lands on its partition key or a GSI partition key, so the query stays within DynamoDB's access patterns. The join runs on your machine, which bounds how many rows it can bring together. OpenSearch SQL lists the rules.
DynamoDB in both
Both clients browse and edit DynamoDB tables, run PartiQL, and manage tables and GSIs. DocKit's table management also covers streams and encryption, which DynoTable leaves to the AWS console. DocKit Community connects to DynamoDB with static access keys. AWS profiles and SSO come with Ultimate.
DynoTable reads every profile in ~/.aws, including SSO, on the Free plan.
Beyond PartiQL, its SQL Workbench runs JOIN, GROUP BY and aggregates over
live tables, which PartiQL cannot express. Smart Tables save a join as a
reusable view.
AI agents and MCP
DocKit Ultimate's agent turns natural language into Elasticsearch DSL, PartiQL or MongoDB queries and can update documents, delete records and create indexes itself, behind per-source permissions and a confirmation for destructive operations. You bring a key for one of ten providers, including Ollama and LM Studio for local models. Amazon Bedrock is not among them.
DynoTable's agent runs on Amazon Bedrock in your own AWS account. It reads index mappings and table schemas, searches indices and runs SQL across both engines, and it can only stage a change. Nothing it writes reaches OpenSearch or DynamoDB until you commit the diff. The built-in MCP server gives the same guarantee to Claude Code, Cursor or Codex. DocKit Ultimate also ships an MCP server.
Price and licence
DocKit Community is free under Apache 2.0. Ultimate costs $99 a year or $9.90 a month, with a 7-day trial, and a release published while your subscription is active stays unlocked after it ends.
DynoTable is closed source. Its Free plan has no time limit and needs a free
account: browsing, PartiQL SELECT, OpenSearch index search, exports and MCP
reads. A paid seat adds editing, the SQL Workbench, Smart Tables and the AI
agent, from $9/mo billed annually ($108/yr), with a 30-day trial and no
credit card. See pricing.
Try DynoTable on the same cluster
Download DynoTable for macOS (Apple Silicon or Intel), Windows or Linux, turn on OpenSearch for your profile, and open an index from ⌘K. Connecting OpenSearch covers domains, Serverless collections and self-hosted clusters.
FAQ
Is DynoTable a DocKit alternative?
Yes, for Amazon OpenSearch and DynamoDB. DynoTable signs OpenSearch requests with your AWS profile, stages document edits as a reviewable diff, and runs SQL that joins an index to a DynamoDB table. DocKit covers more datastores, including Elasticsearch, MongoDB and EasySearch, and its Community edition is free and open source.
Does DocKit support AWS IAM authentication for OpenSearch?
Not in its source as of October 2026. DocKit's Elasticsearch and OpenSearch connection sends a username and password or an API key; AWS request signing exists only in its DynamoDB client, where AWS profile and SSO sign-in is an Ultimate feature. A domain with fine-grained access control and an internal user works; an IAM-only domain or a Serverless collection expects signed requests.
Can DocKit join OpenSearch and DynamoDB data?
No. DocKit keeps each data source in its own editor, and its SQL for OpenSearch goes to the OpenSearch SQL plugin, which joins indices on a best-effort basis and does not aggregate joined results. DynoTable's SQL Workbench joins an index to a DynamoDB table in one statement and runs GROUP BY over the joined rows.
Is DocKit free?
DocKit Community is free and open source under Apache 2.0. The Ultimate edition, which adds the AI agent, the MCP server, cluster monitoring and AWS profile or SSO sign-in, costs $99 a year or $9.90 a month. DynoTable's Free plan has no time limit and includes OpenSearch index search; paid seats start at $9 a month billed annually.
Related
- The OpenSearch Dashboards comparison covers the AWS-bundled default.
- DocKit is also reviewed among DynamoDB clients in the DynamoDB GUI roundup.
- Browse the full comparison hub.
Last verified 2026-10-29 against DocKit v2.0.1. DocKit is a product of GEEKFUN. OpenSearch is a trademark of LF Projects, LLC. AWS, Amazon OpenSearch Service and DynamoDB are trademarks of Amazon Web Services; referenced here for identification only.