Living collection
Open Source Apps That Replace Paid Software
23 resourcesOrganized by use
Why I make these lists
Using AI or having a good setup is no longer enough. Everything changes too fast, and keeping up takes hours of testing, reading and far too many open tabs. I love this stuff. I enjoy discovering tools, trying them and sorting the useful ones from the rubbish. That is why I put these collections together: the things I try and the opinions that actually help people decide. The idea is simple: save you time and give you something you can use. If something is missing or you know a tool that belongs here, tell me.
Signed: Dani
Dani’s favorites
Server · media
Jellyfin
Organize and play your media library from a server you control.
- Own server
- Multiple devices
App · workouts
openGym
How to use this collection
- Use case: start with the problem you need to solve and jump to its category.
- Cost: open source does not always mean free. Check pricing, hosting and maintenance.
- Control: check whether you can host it yourself and where your data lives.
- Limits: read the Watch out section before migrating anything important.
Explore by category
Work and organization
4 resourcesDesign, video and content
5 resourcesData and analytics
3 resourcesCommunication and automation
6 here · 7 totalInfrastructure and privacy
2 resourcesMedia and entertainment
2 resourcesFull profiles and sources
Jellyfin
App
Sources checked: 19 Aug 2026
A self-hosted media server for managing and streaming films, series and music without relying on third-party services
Free, open source media server for organizing and streaming your films, series and music from your own hardware.
Best for: A self-hosted media server for managing and streaming films, series and music without relying on third-party services
Why it belongs here. Organizes and plays your own media library from a server you control, with clients for accessing the files you add.
Its limit. It does not include a film or TV catalog or guarantee smooth playback of every format. Hardware, client, codec and subtitles affect the result.
Your video library on your own server. You supply and organize the files; Jellyfin displays and plays them. Test a file and a client before moving the whole library.
Compatibility and setupRequirements and agent differences
Server and clients. Playback depends on the files, client and server. A file can play directly or need conversion; test the effect of subtitles too.
Acceleration. The official documentation distinguishes supported systems and technologies. Transcoding support does not mean that any computer can play every 4K HDR video smoothly.
Real cost and licenseResource, agent and extra costs
Project license. Jellyfin publishes its server as free software and does not charge a project license fee for use. You still need hardware, storage, electricity, maintenance and your own content.
Example and alternativeWhen it helps and what else to use
First steps. Start on your local network with familiar files. Check whether each client plays directly or converts, review names and folders, and test a restore before expanding the library.
Use in marketing teams. It can provide access to finished pieces or internal material. The review does not present it as a replacement for approvals, frame comments or rights management.
Sources and reviewWhat is documented and what was tested
Verified scope. The Marketing Ultra review, updated on 11 September 2026, examined documentation, video and original opinions. It states that no new installation or playback test was performed for that update.
Resource detailsEditorial profile and links
- Replaces: Plex, Emby
- Deployment: self-hosted, desktop, CLI
- Platforms: Linux, Windows, macOS, Docker, Android, iOS, Apple TV, Amazon Fire TV, Roku, webOS, Tizen
- Watch out: Official apps are not available for every TV platform; some depend on third-party clients such as Swiftfin or Findroid
openGym
App
Sources checked: 2 Sep 2026
Logging workouts with a library of 1,300+ exercises, progress tracking and one-rep max estimates
Self-hosted workout tracking app with more than 1,300 animated exercises, set logging and progress charts, without an account or subscription.
Best for: Logging workouts with a library of 1,300+ exercises, progress tracking and one-rep max estimates
Why it belongs here. Logs routines, sets and progress in an app you can host yourself, with export options documented by the project.
Its limit. The review examined the demo, repository and opinions; it does not establish a self-hosted installation, complete migration or long-term use.
Keep your training history without tying it to a subscription. Before moving all your data, try one routine and make sure you can recover it afterwards.
Compatibility and setupRequirements and agent differences
Ways to use it. The project documents a self-hosted server, PWA and Android APK. Check the current version before choosing a route; the review did not verify a complete installation.
Existing data. The README documents JSON export and import and importers for other apps. Check a sample of your exercises, dates, units and sets before migrating everything.
Real cost and licenseResource, agent and extra costs
Software and server. The app describes itself as free and self-hostable. That choice transfers the cost and maintenance of hardware, backups and updates to whoever installs it.
Example and alternativeWhen it helps and what else to use
First steps. Explore the demo, log a sample workout and test an export. For group use, check isolation between profiles and a restore before adopting it.
Sources and reviewWhat is documented and what was tested
Verified scope. The Marketing Ultra review, updated on 11 September 2026, describes examining the demo, repository and external opinions, without installing a server or migrating training history. This entry preserves that limit.
Resource detailsEditorial profile and links
- Replaces: Strong, JEFIT or other gym logging apps
- Deployment: self-hosted
- Platforms: Web (PWA), Android (APK), iOS (PWA)
- Compatibility: On iOS it can only be installed as a PWA through Safari; there is no native App Store app
- Watch out: Cross-device sync and passkey sign-in require hosting the server with Docker; the browser demo does not include them
Twenty
Resource
Sources checked: 18 Sep 2026
Teams that want a modern CRM and control over their data.
Open source CRM for managing contacts, companies and sales opportunities, with managed cloud hosting or your own server.
Best for: Teams that want a modern CRM and control over their data.
Why it belongs here. It brings companies, contacts and sales opportunities into an interface you can adapt to your business. Choose a managed service or host the CRM on your own infrastructure.
Its limit. Hosting it yourself also means maintaining it: the server, updates and backups. Some organization and security features need a commercial plan. Open source does not make every option free.
If your leads are scattered across spreadsheets, email and notes, Twenty brings them together so you can see which conversation needs a follow-up. Pay for a ready-to-use service or run it yourself, with the technical work that involves.
Compatibility and setupRequirements and agent differences
How to use it In your browser, through Twenty Cloud or your own installation. The cloud service takes server administration off your hands.
Self-hosting The official documentation provides Docker Compose installation instructions and asks for at least 2 GB of RAM. You need to configure the public address, HTTPS, persistent storage and backups. That minimum does not establish capacity for every team size.
Real cost and licenseResource, agent and extra costs
Software and infrastructure The self-hosted Pro plan has a license cost of €0 per user per month. You pay for the server and its maintenance; commercial features are priced separately.
Managed service Twenty Cloud Pro works out at roughly €10.47 per user per month on monthly billing, or €7.85 per month on annual billing. Organization is about €21.82 per month, or €16.58 with annual billing. These indicative conversions use the ECB exchange rate from 18 September 2026; the provider lists prices in dollars. Taxes and the exchange rate at payment affect the final amount.
License The project combines code mainly under AGPLv3, MIT packages and commercially licensed files. Check the license for the features you plan to use, especially if you modify or redistribute the product.
Example and alternativeWhen it helps and what else to use
Example use case A small team can record an interested company, attach a contact, open a sales opportunity and set the next task. This is an example to guide a trial, not Dani's personal experience or a performance test.
What to compare If you already use HubSpot, Pipedrive or Salesforce, check which processes, automations and integrations you need to retain before considering a move. A simpler interface does not guarantee that Twenty covers your current setup.
Sources and reviewWhat is documented and what was tested
Review date Sources consulted on 18 September 2026. Prices checked against the billing and hosting selector on the official website.
Scope This entry relies on documentation and public evidence. The demo.twenty.com address redirected to Twenty Cloud's sign-in page; no account was created and no migration with real data was tested.
Product ↗Official pricing ↗ECB reference exchange rate ↗Docker Compose installation ↗Code and license ↗Terms and hosting options ↗
Resource detailsEditorial profile and links
- Type: Web CRM.
- Deployment: Cloud or your own server.
- License: Mainly AGPLv3, with MIT packages and commercial components.
- Base self-hosted license: €0; infrastructure and maintenance cost extra.
LifeForge
App
Sources checked: 6 Sep 2026
Bringing tasks, finances, notes, projects and photos into one self-hosted dashboard without relying on external SaaS tools
Modular self-hosted dashboard that brings tasks, finances, a journal, notes and photos into one web app, with your data under your control.
Best for: Bringing tasks, finances, notes, projects and photos into one self-hosted dashboard without relying on external SaaS tools
Resource detailsEditorial profile and links
- Replaces: Notion + YNAB + Google Photos, for people who prefer self-hosting
- Deployment: self-hosted, with Docker or manual installation
- Platforms: Web browser
- Watch out: Early development: modules and architecture can change; the CC BY-NC-SA 4.0 license prohibits commercial use
HistApp
License: MIT
Sources checked: 13 Sep 2026
Legacy app recovered from the Inbox.
Resource detailsEditorial profile and links
OpenCut
App
Sources checked: 17 Aug 2026
Creators who want timeline editing without depending on ByteDance.
Open source video editor built as a free alternative to CapCut, with timeline editing for creators who want to avoid depending on ByteDance.
Best for: Creators who want timeline editing without depending on ByteDance.
Why it belongs here. It offers an open alternative to CapCut; the classic editor lets you explore the current workflow while the team develops a rewrite.
Its limit. The rewrite announces future features. The current review did not repeat an export test, so it does not establish that OpenCut is ready for client work.
Try the existing editor, but before using it for a job, import your files, edit, save and open the exported video outside the editor. Promised features in the new version do not count as available.
Compatibility and setupRequirements and agent differences
Two different versions. The official project separates the classic editor available at opencut.app from the rewrite in development. Do not apply screenshots or promises from one version to the other.
Before delivery. Check input formats, cutting accuracy, project saving and repeatable export in the version and environment you intend to use.
Real cost and licenseResource, agent and extra costs
Open source. The project publishes open source code, but this entry does not establish savings over CapCut. The practical cost includes testing time, formats and corrections when a workflow fails to produce a usable file.
Example and alternativeWhen it helps and what else to use
Minimum test. Start by editing a short clip with material you can share. Then repeat with a project similar to your real work and open the exported file outside the editor.
If you need to deliver now. The review recommends comparing established editors such as Shotcut and Kdenlive. They are editing alternatives, not identical CapCut clones.
Sources and reviewWhat is documented and what was tested
Verified scope. The Marketing Ultra review, checked on 11 September 2026, examined both repositories and an external demonstration. It explicitly states that no new export test was performed.
Resource detailsEditorial profile and links
- Replaces: CapCut.
DramaClaw
App
Turning scripts or novels into finished videos through an automated end-to-end pipeline
Self-hosted tool that coordinates scripts, voices and AI video editing for creators and small studios.
Why it belongs here. It brings scripts, planning, visual assets, voices and editing into a self-hosted system with a node canvas and an episode-based production chain.
Its limit. Elastic License 2.0 restricts offering DramaClaw as a hosted service. The maintainer also says that a small Powered by DramaClaw credit must remain in the interface. External generation services have their own costs.
In plain English: bring in a story, organize the characters and scenes, connect your AI models and manage everything from the storyboard to the finished video without jumping between a dozen programs.
Compatibility and setupRequirements and agent differences
Compatibility. Runs through Docker on macOS, Windows with Docker Desktop and WSL2, and Linux. The documentation recommends 2 vCPUs and 4 GB of RAM; a GPU is only required for the optional 3D world feature.
Real cost and licenseResource, agent and extra costs
Cost. The code is published under Elastic License 2.0. Separately, the maintainer says that a small Powered by DramaClaw credit must remain in the interface. RelayClaw calls and other providers have their own costs; this review does not establish verified euro prices.
Example and alternativeWhen it helps and what else to use
Example. A small studio can keep scripts, characters, settings, initial images, voices and editing in the same project to produce a series of related videos.
Alternative. A manual workflow with separate tools for text, images, voice, video and editing gives you more freedom, but leaves you coordinating files, providers and context yourself.
Sources and reviewWhat is documented and what was tested
Review. Entry checked on 27 September 2026 against the repository and supplied primary documentation. No personal testing, Dani rating or social consensus is claimed.
Resource detailsEditorial profile and links
- Replaces: Manually assembling separate AI tools for images, voice and video at each production stage
- Deployment: self-hosted, web
- Platforms: Linux, macOS, Windows (Docker Desktop + WSL2)
- Watch out: Model inference runs remotely through your own gateway or a third-party one; the local machine only orchestrates the pipeline. The 3D features require a CUDA GPU.
Jaaz
App
Sources checked: 5 Sep 2026
Generating AI images and video from a visual canvas without complex prompts, with local execution
Open source creative assistant with an infinite canvas where you draw, point and chat to generate AI images and video, with local execution for privacy.
Best for: Generating AI images and video from a visual canvas without complex prompts, with local execution
Resource detailsEditorial profile and links
- Replaces: Canva AI
- Deployment: desktop, self-hosted
- Platforms: Windows, macOS, Linux
- Watch out: The community license does not allow code modification, redistribution or deployment for simultaneous team use; all of those require a commercial license
CutScript
App
Sources checked: 21 Sep 2026
Editing talking-head videos, podcasts and tutorials by cutting and rearranging clips through the transcript
Free local video editor that turns the transcript into your timeline: delete a word from the text and it disappears from the video.
Why it belongs here. It turns the transcript into an editing surface and brings editing, subtitles, audio cleanup and clip creation into one local application under the MIT license.
Its limit. The documented installation is technical, and the dossier contains no independent testing of accuracy, stability, speed or handling of long projects.
In plain English: load a video, let CutScript transcribe the speech and trim it by deleting words from the text.
Compatibility and setupRequirements and agent differences
Compatibility. Electron application for Windows and macOS. Requires Node.js 18+, Python 3.10+ and FFmpeg on the PATH; Ollama is optional.
Real cost and licenseResource, agent and extra costs
Cost. The repository uses the MIT license and describes itself as open source. Optional OpenAI or Claude connections have external costs, but the dossier supplies no verified official pricing in euros.
Example and alternativeWhen it helps and what else to use
Example. Edit an interview through its transcript, remove passages and export the video or subtitles in SRT, VTT and ASS formats.
Alternative. Descript is the only verified alternative in this entry, although there is no direct comparison that establishes a winner.
Sources and reviewWhat is documented and what was tested
Review. Entry and documentation consulted on 29 September 2026; the latest repository activity supplied is from 6 March 2026.
Resource detailsEditorial profile and links
- Replaces: Descript
- Deployment: desktop
- Platforms: Windows, macOS, Linux
- Compatibility: Requires Node.js 18+, Python 3.10+ and FFmpeg; a GPU is recommended for WhisperX transcription
- Watch out: Only covers text-based editorial editing; it does not replace VFX, color grading or motion graphics tools
Shrimply
App
Sources checked: 11 Sep 2026
Video editing with visual effects, keyframe animation, 3D scenes and optional AI features
Open source video editor written in Rust, with keyframes, 3D scenes, audio processing, lip sync and an optional AI server for transcription and video generation.
Why it belongs here. The only open source video editor described here that combines non-linear editing, local AI for transcription, video generation and lip sync, and editor automation through MCP from coding agents.
Its limit. Pre-alpha: the README itself warns about random crashes, irreversible project loss and resource leaks. Unsuitable for production.
A free video editing program with AI features such as automatic subtitles, video generation and lip sync, without having to send your files online. Everything runs on your computer.
Compatibility and setupRequirements and agent differences
Compatibility. Cross-platform according to the project, with documented Flatpak installation on Linux. GTK 4 and libadwaita interface. AI features require an NVIDIA GPU with CUDA.
Real cost and licenseResource, agent and extra costs
Cost. Free, under GPL-3.0. The AI features depend on separately licensed proprietary NVIDIA components: CUDA Toolkit, OptiX and Video Codec SDK.
Example and alternativeWhen it helps and what else to use
Example. Install through Flatpak, configure the Python compute server for AI, import a video and use automatic transcription to generate subtitles inside the editor.
Alternative. Kdenlive or Shotcut for stable editing without integrated AI; Whisper or other separate transcription tools for AI functions.
Sources and reviewWhat is documented and what was tested
Review. Review based on the official documentation and repository README. The project is pre-alpha, with no visible community or documented user opinions.
Resource detailsEditorial profile and links
- Replaces: Kdenlive, Shotcut or other desktop video editors
- Deployment: desktop
- Platforms: Linux
- Compatibility: Some AI features require proprietary NVIDIA components, including CUDA Toolkit and OptiX SDK
- Watch out: Pre-alpha: the repository warns about crashes, performance regressions and irreversible corruption of project files
Public APIs / Datasets repository
Library
Finding free public APIs organized by category for integration into software projects
Community directory of hundreds of free public APIs organized by category, for developers looking for data sources for their projects.
Why it belongs here. It gathers hundreds of public APIs by category and lets you shortlist candidates by authentication, HTTPS and CORS before checking each provider.
Its limit. Quotas, stability, pricing and production readiness are not documented consistently; the repository's MIT license does not cover the terms of the linked APIs.
In plain English: a huge, organized list for finding an API that supplies data to a project. You still need to check whether that API is reliable and what it costs.
Compatibility and setupRequirements and agent differences
Compatibility. The repository classifies entries by authentication, HTTPS and CORS. Requirements vary by provider; a shared account and key apply only to APILayer's own suite.
Real cost and licenseResource, agent and extra costs
Cost. The repository is open source under the MIT license. Each API has its own pricing and limits; the dossier supplies no euro prices or verifiable official conversion.
Example and alternativeWhen it helps and what else to use
Example. Choose an API listed as requiring no authentication and supporting HTTPS and CORS, confirm those details in its documentation and make a minimal request before integrating it.
Alternative. Search Google for public APIs manually and check each provider's documentation directly.
Sources and reviewWhat is documented and what was tested
Review. Entry checked on 27 September 2026 against the repository, README and APILayer website; no personal test by Dani is recorded.
Official repository ↗APILayer website ↗Official Postman collection ↗
Resource detailsEditorial profile and links
- Replaces: manually searching Google for public APIs
- Deployment: web
- Watch out: A static curated Markdown list; entries can be outdated or contain broken links because it relies on community contributions
Graphic Walker
Library
Sources checked: 21 Sep 2026
Embedding interactive Tableau-style data visualization in your own apps without depending on heavy BI platforms
Open source visual analytics component with drag-and-drop controls, embedded in web apps or Python notebooks as a lightweight alternative to Tableau.
Why it belongs here. It adds drag-and-drop visual exploration to a React application without requiring a monolithic business intelligence platform from the outset.
Its limit. Real integration requires development, data preparation, permissions and your own Spanish translation. The dossier does not establish performance at large data volumes or supply enterprise prices officially converted to euros.
In plain English: feed it data and build charts by dragging columns with your mouse, either in the web version or inside your own React application.
Compatibility and setupRequirements and agent differences
Compatibility. Works on the web, as an embeddable React component and in self-hosted installations. It can compute in the browser through Web Workers or delegate queries to external infrastructure; Spanish requires custom i18n resources.
Real cost and licenseResource, agent and extra costs
Cost. The project uses Apache-2.0 and lists no fee for the open source package. Hosting, development and enterprise services cost extra; no price is published here without an official, dated euro conversion.
Example and alternativeWhen it helps and what else to use
Example. Embed a view in an internal application to explore spend and conversions by channel and date. This is a proposed use case, not a personal test.
Alternative. Tableau is the stated alternative, but it offers a broader platform. Compare connectors, governance and collaboration before treating the two as interchangeable.
Sources and reviewWhat is documented and what was tested
Review. Entry checked on 2 October 2026 against the repository, official website and primary documentation; no personal test or rating is attributed to Dani.
Resource detailsEditorial profile and links
- Replaces: Tableau
- Deployment: web, extension
- Platforms: React, JavaScript/TypeScript, Python (PyGWalker)
- Watch out: An embeddable component, not a complete BI platform: data source and user management have to be integrated separately
ChartDB
App
Sources checked: 6 Sep 2026
Visualizing and designing database schemas by importing them with a single query
Database diagram editor that imports your schema with a query, lets you edit it visually, export DDL or migrate between SQL dialects with AI.
Best for: Visualizing and designing database schemas by importing them with a single query
Resource detailsEditorial profile and links
- Replaces: DrawSQL, dbdiagram.io, SQLDBM
- Deployment: web, self-hosted
- Platforms: PostgreSQL, MySQL, SQL Server, SQLite, MariaDB, ClickHouse, CockroachDB, Oracle, Snowflake, BigQuery
- Watch out: AI export features require an OpenAI key or your own inference server
OpenWA
Library
Bots, notifications and automations through WhatsApp Web.
Node.js library that turns WhatsApp Web into a programmable API for bots, notifications and automations, with self-hosting and a paid plan for advanced features.
Why it belongs here. It turns WhatsApp Web into a programmable API. The v5 alpha includes an MCP server for connecting AI agents.
Its limit. Unofficial project: the repository warns that use is at your own risk. V5 is in alpha; the documented stable version is 4.76.0.
Run an instance, link WhatsApp through a QR code and use its interface to send messages and receive events. The core is free; some restricted features require a license key.
Compatibility and setupRequirements and agent differences
Compatibility. The documented stable version is 4.76.0 and runs on Node.js. The reviewed v5 alpha instructions describe interchangeable drivers and an MCP server; check before applying them to the stable branch.
Real cost and licenseResource, agent and extra costs
Cost. The core is free under the Hippocratic + Do Not Harm license. Some restricted features require a license key; check the terms published at openwa.dev before planning a deployment.
Example and alternativeWhen it helps and what else to use
Example. Launch the server with npx @open-wa/wa-automate@4.76.0 --port 8080, scan the QR code with your phone and send messages, receive webhooks or connect a bot from another application.
Alternative. The official WhatsApp Business platform or SaaS products that offer WhatsApp as a channel. Check each option's current requirements and terms.
Sources and reviewWhat is documented and what was tested
Review. Based on the official documentation, repository README and verified Hacker News comments. No personal testing by the Marketing Ultra team is included.
Resource detailsEditorial profile and links
- Deployment: Node.js and self-hosted.
Meetily
App
Transcribing and summarizing meetings locally without sending data to the cloud
Open source meeting assistant that records, transcribes with Whisper and summarizes with local AI. Everything stays on your computer, without cloud storage or bots joining the call.
Why it belongs here. Meetily keeps recordings and transcripts on the user's computer. Summaries can also be generated there with a local model such as Ollama; choosing an external provider sends the transcript to that service. The Community edition is published under MIT.
Its limit. Linux requires building from source, and several advanced features belong to Pro. The dossier also contains no independent accuracy test.
In plain English: it listens to a meeting on your computer and writes down what is said without forcing you to store the audio in the cloud. The summary stays on the computer only when you use a local model.
Compatibility and setupRequirements and agent differences
Compatibility. Packages are available for macOS and Windows. Linux requires building from source; it works with meeting apps by capturing microphone and system audio.
Requirements. Published minimum requirements are 8 GB of RAM, a four-core CPU, 10 GB of storage, Node.js 18+, Python 3.10+ and FFmpeg; a GPU is optional.
Real cost and licenseResource, agent and extra costs
Cost. The Community edition is free under MIT. A Pro edition exists, but no price is published here because the dossier lacks a complete, official and dated euro conversion.
Example and alternativeWhen it helps and what else to use
Example. Use it to transcribe a client interview and then generate a local summary with Ollama, with the appropriate consent.
Alternative. The entry identifies Otter.ai and Fireflies.ai as replacements; the dossier lacks sufficient independent comparison to name an overall winner.
Sources and reviewWhat is documented and what was tested
Review. Information checked against the official website, repository, installation documentation and Meetily Pro page, reviewed in the dossier on 25 September 2026.
Official website ↗Official repository ↗Releases ↗Meetily Pro ↗
Resource detailsEditorial profile and links
- Replaces: Otter.ai, Fireflies.ai
- Deployment: desktop, self-hosted
- Platforms: macOS, Windows, Linux
- Compatibility: No prebuilt Linux installer; building from source is required
- Watch out: Requires at least 8 GB of RAM and a compatible GPU for acceptable transcription performance
OpenDroid
App
Automating everyday Android tasks through voice or text commands with an autonomous AI agent
Autonomous open source AI agent for Android that plans, executes and checks tasks on your phone using more than 60 actions and 12 LLM providers.
Why it belongs here. OpenDroid combines autonomous planning, more than 60 Android actions, memory and support for 12 LLM providers in an open source project.
Its limit. It requires highly sensitive permissions, and the dossier includes no independent test confirming the reliability of all advertised automations.
In plain English: give your phone an instruction and OpenDroid attempts to break it into steps, run them across apps and check whether they worked.
Compatibility and setupRequirements and agent differences
Compatibility. Native Android application. Building it requires JDK 21 and Android SDK 35; it uses cloud or local models through supported providers.
Real cost and licenseResource, agent and extra costs
Cost. OpenDroid describes itself as free and open source. APIs, voices and infrastructure have separate costs, but the dossier supplies no verified official euro prices.
Example and alternativeWhen it helps and what else to use
Example. It can plan a routine that looks up information, creates an alarm and prepares a message, once the required permissions have been granted.
Alternative. Tasker and MacroDroid are the alternatives identified in this entry, although the dossier does not support a rigorous price or feature comparison.
Sources and reviewWhat is documented and what was tested
Review. Information checked against the repository, README and official website on 27 September 2026. No independent tests or verified social opinions were supplied.
Resource detailsEditorial profile and links
- Replaces: Tasker, MacroDroid
- Deployment: desktop (Android APK)
- Platforms: Android
- Compatibility: Requires Android 15, SDK 35, and enabled accessibility permissions
- Watch out: Building requires JDK 21, and accessibility permissions can trigger system security warnings
InteraOne
App
Sources checked: 21 Sep 2026
Customer support with AI and human agents in one inbox, as an open source alternative to Intercom and Zendesk
Open source customer support platform that combines AI and human agents in a single inbox, deployed through Docker or the cloud.
Why it belongs here. It brings AI and human agents, tickets, contacts and multiple channels into an open source platform you can host on your own infrastructure.
Its limit. The dossier contains no independent performance or stability test, and verified total costs officially converted to euros are missing.
In plain English: it gathers questions from your website, email, WhatsApp and Telegram so an AI can respond and a person can pick up the conversation from the same inbox.
Compatibility and setupRequirements and agent differences
Compatibility. Runs as a web service or through your own Docker deployment; it offers an embeddable widget and stated connections to email, WhatsApp and Telegram.
Infrastructure. The stated architecture uses MongoDB, Redis, Qdrant and a compatible LLM provider from OpenAI, Google, AWS Bedrock or Hugging Face. The repository documents starting a development environment, not a production deployment procedure.
Real cost and licenseResource, agent and extra costs
Cost. The core is distributed under Apache-2.0, with an enterprise edition under its own license in /ee. The dossier lists a free Starter plan and two paid plans, but their prices are only in dollars and do not establish the full euro cost of infrastructure, models and maintenance.
Example and alternativeWhen it helps and what else to use
Example. A digital product can bring widget, email, WhatsApp and Telegram queries into one inbox, letting a person take over when needed.
Alternative. Intercom and Zendesk are the listed replacements, but there is no current like-for-like comparison in euros that establishes a winner.
Sources and reviewWhat is documented and what was tested
Review. Entry checked on 1 October 2026 against the official website and repository. No independent tests, verified public opinions or commercial case studies were received.
Resource detailsEditorial profile and links
- Replaces: Intercom, Zendesk
- Deployment: web, self-hosted with Docker
- Platforms: Web widget, Email, WhatsApp, Telegram
- Watch out: The free plan is limited to 50 AI messages per month and one human agent; Enterprise in /ee is not Apache 2.0
OpenWhispr
App
Sources checked: 5 Sep 2026
Voice dictation in any desktop app with private, local transcription
Open source dictation app that turns speech into text in any application: transcribes locally with Whisper or Parakeet, records meetings with speaker identification and organizes notes with semantic search.
Best for: Voice dictation in any desktop app with private, local transcription
Resource detailsEditorial profile and links
- Replaces: Wispr Flow, Aqua Voice
- Deployment: desktop
- Platforms: macOS, Windows, Linux, iOS
- Compatibility: Speaker identification is unavailable on Intel Macs because of ONNX Runtime incompatibility
- Watch out: Cloud transcription requires the user's own API keys, or BYOK
OpenCLI
App
Sources checked: 6 Sep 2026
Turning any website, Electron app or local binary into CLI commands that people and AI agents can run
Turns websites, Electron apps and local tools into one CLI interface so AI agents can operate your browser through existing authenticated sessions.
Best for: Turning any website, Electron app or local binary into CLI commands that people and AI agents can run
Resource detailsEditorial profile and links
- Replaces: Manual scraping or automation scripts using Puppeteer or Playwright
- Deployment: CLI, extension
- Platforms: Node.js, Chrome, Claude Code, Cursor
- Watch out: Requires Node.js >= 20.18.1 and an installed Chrome extension for browser automation
Awesome LLM Apps
Library
Starting AI projects from runnable examples of RAG, agents and chatbots
Collection of more than 100 Python applications with runnable code for RAG, agents and chatbots across providers, for learning by building.
Why it belongs here. It brings together more than 100 open source resources spanning agents, skills and RAG applications, with memory, MCP and multi-agent team examples under Apache-2.0.
Its limit. Recipes are not production-ready products: each project requires checking dependencies, keys, data, security and external costs.
In plain English: choose an application close to your idea, run it and modify its code instead of starting from an empty folder.
Compatibility and setupRequirements and agent differences
Compatibility. Install locally by cloning the repository. Requires Python, pip and keys for the model or service used. Includes examples for Claude, Gemini, GPT, DeepSeek, Llama, Qwen, Streamlit, MCP and several agent frameworks.
Real cost and licenseResource, agent and extra costs
Cost. The code costs €0 and is published under Apache-2.0. APIs, hosting and connected services have their own costs, but the dossier supplies no verified prices to quantify them.
Example and alternativeWhen it helps and what else to use
Example. Clone the repository, open starter_ai_agents/ai_travel_agent, install requirements.txt and run travel_agent.py with Streamlit.
Alternative. Following scattered tutorials or building your own starter structure gives you more control, but leaves you responsible for architecture and integration from scratch.
Sources and reviewWhat is documented and what was tested
Review. Information checked on 23 September 2026 against the official introductory file and repository details; no personal tests or production results are claimed.
Resource detailsEditorial profile and links
- Replaces: scattered tutorials and your own boilerplate for LLM apps
- Deployment: self-hosted
- Platforms: Python, Streamlit, OpenAI, Anthropic, Google Gemini, Ollama, LangChain, LlamaIndex, CrewAI
- Watch out: Most apps require API keys from paid providers such as OpenAI, Anthropic and Google; only some work with local models through Ollama
Trek
App
Sources checked: 6 Sep 2026
Planning group trips with real-time collaborative itineraries, shared budgets and interactive maps
Self-hosted travel planner with real-time collaboration, maps, budgets, packing lists, a journal and an integrated AI assistant.
Best for: Planning group trips with real-time collaborative itineraries, shared budgets and interactive maps
Resource detailsEditorial profile and links
- Replaces: Wanderlog, TripIt
- Deployment: self-hosted, web
- Platforms: Docker, Docker Compose, Kubernetes/Helm, Unraid, Proxmox, Raspberry Pi, PWA (iOS/Android)
- Watch out: Full place search requires a Google Places API key; without one it uses OpenStreetMap with more limited results
OmniVoice Studio
App
Sources checked: 5 Sep 2026
Local voice cloning and multilingual video dubbing without relying on paid APIs
Desktop app that clones voices, dubs videos and creates audiobooks in 646 languages, running entirely locally without usage limits or API keys.
Best for: Local voice cloning and multilingual video dubbing without relying on paid APIs
Resource detailsEditorial profile and links
- Replaces: ElevenLabs
- Deployment: desktop, self-hosted
- Platforms: macOS (Apple Silicon), Windows, Linux, Docker
- Compatibility: Intel Macs are not supported by the local AI backend
- Watch out: In active beta; the default voice engine's weights use CC-BY-NC, restricting commercial use unless you change engines
How to read the entries
Sources checked means the documentation was reviewed on the stated date. If a tool has been tested or has a full analysis, you will find the link. Fields without evidence stay empty. Better a gap than a made-up certainty.
Frequently asked questions about open source apps
Does open source mean free?
No. The code can be open while a hosted plan, infrastructure or commercial features cost money. Each entry separates the license from the price so the two do not get confused.
Can every app here be self-hosted?
Not necessarily. Some offer your own installation, others prioritize a managed web version, and others require technical knowledge. The entry states the deployment type when it has been confirmed.
Are these all marketing tools?
No. This is Marketing Ultra's general collection of open source apps. Marketing and AI feature heavily because those are the areas we work in most, alongside productivity, content, data, communication and infrastructure.