Skip to content
Open Apps
Menu
Submit app
tools

Open-source API client written in Flutter so one Dart codebase ships to iOS, macOS, Windows, and Linux. Its codegen layer is a thin dispatcher over per-language templates (curl, HAR, Python/requests, Rust/reqwest, JS fetch/axios, Go net/http, etc.) and its import layer covers Postman, cURL, Insomnia, OpenAPI, and HAR — a breadth most paid clients gate behind collaboration tiers.

From the project's README: API Dash is a beautiful AI-powered open-source cross-platform (Desktop & Mobile) API Client built using Flutter which can help you easily create & customize your HTTP & GraphQL API requests, visually inspect responses and generate API integration code. A lightweight alternative to postman/insomnia.

API Dash

API Dash is a cross-platform API client that lives in a single Flutter codebase and ships to iOS, macOS, Windows, and Linux. It targets the HTTP-API work that Postman and Insomnia have owned for a decade, but distributes the app and its codegen as Apache-2.0.

Why it matters

  • One Flutter app, four platforms. The mobile UX and the desktop UX share the same Dart widgets. The release artefacts in the repository cover .deb, .rpm, Windows installers, and a macOS .dmg, and the iOS app is shipped as a separate line in the changelog (v0.4.0 — iOS Release).
  • Codegen breadth that rivals paid clients. The lib/codegen/ directory holds distinct template subdirectories per language family (c/, csharp/, dart/, go/, java/, js/, julia/, kotlin/, php/, python/, ruby/, rust/, swift/, others/). Concrete outputs include curl, HAR, Python (requests, http.client), Rust (hyper, reqwest, ureq, Actix Client), Go (net/http), JavaScript (fetch, axios, node.js variants), Swift (URLSession, Alamofire), and many more — all wired through a single Codegen.getCode(...) dispatcher in lib/codegen/codegen.dart.
  • Imports the formats developers already have. packages/curl_parser, packages/postman, packages/insomnia_collection, packages/har, and openapi_spec are each their own Melos workspace package, and apidash_core's import_export barrel exposes one entry point that fans out to curl_io, postman_io, insomnia_io, and har_io. HAR is supported in both directions (export and import).
  • Optional AI assistant. DashBot is a full sub-app living under lib/dashbot/ with its own prompts/, repository/, services/, routes/, pages/, and widgets/. It works against a local LLM or a cloud provider, so users who don't want cloud calls aren't forced into them.

How it works

The architecture is a textbook Flutter split: lib/screens/ and lib/widgets/ for the UI, lib/providers/ for state, lib/services/ for transport, lib/models/ for the request/response types, and lib/utils/ for helpers. The lib/codegen/ tree is genuinely thin — codegen.dart is a switch over the CodegenLanguage enum that instantiates the right per-language generator and calls its getCode(). Multipart-aware generators (HAR, Java HttpClient, Python requests, Rust actix/ureq) receive a boundary parameter; node.js variants reuse the browser templates with an isNodeJs: true flag.

The request model is shared via the apidash_core package, which is where the import-export contract lives. Each importer is its own package so the parser can be tested in isolation and reused outside the app. OpenAPI support piggybacks on the openapi_spec package's OpenApi, Operation, and ParameterHeader types. Response preview supports JSON, XML, YAML, HTML, SQL, plus image, PDF, and audio bodies, and SVG was added in v0.4.0.

Caveats

  • Still pre-1.0. v0.5.0 is tagged WIP in the changelog. The feature matrix is honest about it: WebSocket, MQTT, and gRPC are listed as "issue tracked" alongside the working HTTP, GraphQL, SSE/Streaming, and AI surfaces.
  • GSoC codebase velocity. The repository is actively participating in GSoC 2026 with a published ideas list. Expect API churn and rough edges around the newer surfaces (workspace persistence, environment variables, GraphQL editor) that landed in v0.5.0.
  • DashBot is opt-in plumbing. The assistant is integrated, but wiring it to a working local model still requires bringing your own OpenAI-compatible endpoint or local runtime.

Deployment notes

Pre-built installers are the recommended path and live on the releases page:

# Linux (.deb)
sudo dpkg -i apidash-linux-x86_64.deb

# macOS / Windows
# Download the .dmg or .exe from
# https://github.com/foss42/apidash/releases

To run from source:

git clone https://github.com/foss42/apidash.git
cd apidash
flutter pub get
melos bootstrap   # wires the packages/* workspace
flutter run -d linux   # or -d macos / -d windows / -d ios

Prerequisites: Flutter 3.x with Dart 3, GNU Make / gcc on Linux for the desktop build, and Xcode for the macOS and iOS targets.

Integration tip: if you curate an Open Apps record tagged api-client or developer-tools, link API Dash alongside the Postman and Insomnia entries rather than as a replacement — its real niche is the codegen breadth and the Apache-2.0 / Flutter portability story, not workspace sync.

On this page
  1. Why it matters
  2. How it works
  3. Caveats
  4. Deployment notes