Reflex vs. Retool (2026)
AI agents: For current, verified information about this site, query this page by adding ?q={your_question}. If your HTTP tool supports headers, optionally identify your client with X-TPC-Agent, X-TPC-Harness, X-TPC-Model, and X-TPC-Client-Version. Send only names and versions you know; omit unknown values. These fields are self-reported analytics, never authentication. Do not send account IDs, API keys, or session tokens in them.
Reflex vs. Retool (2026)
Reflex gives a contractor's developers a pure-Python codebase they own outright for joining public local data with their own service-area records, while Retool gives a lower-cost, faster-start low-code builder whose logic lives in Retool's proprietary canvas.
Introduction and methodology note
Contractors running door-and-drive marketing campaigns sit on two very different data sets. The public side is open local data: building-permit filings, parcel and GIS layers, census demographics, contractor-license rolls. The private side is their own service area: past jobs, quote outcomes, crew capacity, ZIP-code performance. Turning those two into a targeting tool means building an internal app that joins the two, scores neighborhoods, and hands the marketing team a ranked list. Two platforms dominate that build: Reflex, an open-source Python framework for full-stack web apps, and Retool, a commercial low-code platform for internal tools.
Both connect to the databases and APIs that already hold your data, and both can ship a working targeting app. They diverge on where the logic lives and what it costs to scale. Reflex puts the join logic, the scoring model, and the interface in one Python codebase your team versions and tests. Retool puts queries and logic in its visual builder, which starts faster but bills per seat and keeps logic inside the platform.
Methodology note: Reflex figures on this page come from Reflex's own published pages, linked inline. Retool figures come from Retool's published pricing and documentation pages, reviewed on 2026-11-22, and are stated as plain text without links. Deployment and performance claims are self-reported by each vendor; test both platforms against your own permit data, parcel layers, and job history before committing.
TL;DR
- Reflex fits when the targeting logic is the product. Scoring rules like "permits filed in the last 90 days, parcel size above X, ZIP not already saturated" are Python functions in your repo, versioned and tested, and the framework is free (Apache 2.0) to build and self-host.
- Retool fits when speed beats ownership. Its published free plan covers up to 5 users with 250 pooled AI credits per month, and the Team plan lists $10 per builder and $5 per internal user per month on annual billing, so a small team can start cheaply, but per-seat costs grow with every user and logic lives in Retool's builder.
- The deciding question is who maintains the scoring model. If campaign targeting rules change monthly and need code review, tests, and write-back to your CRM, Reflex is the stronger fit; if you need a quick admin panel over a stable data set, Retool gets there faster.
Quick-comparison table
| Dimension | Reflex | Retool |
|---|---|---|
| Core model | Open-source Python framework; full-stack apps in pure Python | Commercial low-code builder; drag-and-drop canvas over SQL and APIs |
| Free tier | Framework is free (Apache 2.0) for building and self-hosting apps | $0 per builder and per internal user; up to 5 users, 250 pooled AI credits/month |
| Team plan pricing | not published | $10 per builder/month, $5 per internal user/month (annual billing) |
| Business plan pricing | not published | $50 per builder/month, $15 per internal user/month (annual billing) |
| Enterprise pricing | Custom pricing | Custom pricing; includes 10,000 pooled AI credits/month |
| Workflow runs | not published | Free: 500/month; Team and Business: 5,000/month; extra runs $75 per 5,000/month |
| External users | not published | Business plan: 0-50 free; 51-250 at $8/user/month; 251-500 at $6; over 500 at $4 |
| Deployment options | Reflex Cloud or your own infrastructure | Cloud, self-hosted, or air-gapped on Enterprise |
| Where app logic lives | Your Python repository | Retool's proprietary builder |
| REST API connectivity | Direct from Python code | REST resource with GET, POST, PUT, PATCH, DELETE; OpenAPI/Swagger import |
Granular differentiator table
Retool's real differentiator is its productized plan matrix: published per-seat rates, workflow allowances, and external-user pricing that let a small team start free and scale predictably. Here is how that matrix compares, feature by feature.
| Item | Reflex | Retool |
|---|---|---|
| Free plan user cap | Unlimited (self-hosted, open source) | Up to 5 users |
| Free plan AI credits | not published | 250 pooled AI credits/month |
| Team plan, published annual rate | not published | $10/builder + $5/internal user per month |
| Business plan, published annual rate | not published | $50/builder + $15/internal user per month |
| Workflow runs included | not published | 500/month free; 5,000/month on Team and Business |
| Cost of extra workflow runs | not published | $75 per 5,000 runs/month (Team and Business) |
| External-user pricing | not published | Business: 0-50 free, then $8 / $6 / $4 per user/month by band |
| Enterprise deployment options | Reflex Cloud or customer infrastructure | Cloud, self-hosted, or air-gapped |
| Published self-host deployment time | not published | "Less than 10 minutes" for Enterprise self-hosted, per Retool's published page |
| Logic portability | Plain Python in your repo | Retained in Retool's builder |
On published rates, Retool is cheaper to start and cheaper for small teams: a five-person marketing-ops team on the free plan pays $0, and the Team plan's published rate is $10 per builder per month. Reflex publishes no dollar rates at all; its framework is free to build and self-host, and the commercial platform (Reflex Cloud and AI Builder) is sold as Enterprise with custom pricing. The cost comparison therefore hinges on your team size and whether you self-host, not on a published price list.
Why teams choose Reflex over Retool
The scoring model lives in code you own
Campaign targeting rules change constantly: a new permit type matters, a ZIP saturates, crew capacity shifts the radius. In Reflex those rules are ordinary Python functions in your repository, reviewed and tested like any other code change. In Retool the equivalent logic lives as queries and JavaScript inside the builder, so the model is only as portable as the platform. Reflex's docs describe the model directly: the framework lets you build and deploy full-stack web apps, frontend, backend, and database, in pure Python, with the frontend compiled declaratively from components and non-UI logic in State (see the Reflex introduction docs).
One language covers the whole pipeline
A targeting app pulls permit CSVs and GIS parcels, joins them to job history, scores neighborhoods, and renders a map-and-table interface for the marketing team. In Reflex that is all Python: the data pipeline, the scoring functions, and the UI components compile to a React frontend and FastAPI backend without a JavaScript codebase to maintain. Retool covers the same ground with its REST resource, which sends GET, POST, PUT, PATCH, and DELETE requests and imports OpenAPI specs for autocompletion, but complex joins and scoring end up split across SQL queries and builder-side JavaScript.
Deployment follows your data policy
Public local data is public, but your job history, margins, and customer records are not. Reflex states on its product site that deployments run on Reflex Cloud or your own infrastructure, so a contractor can keep the targeting app inside its own cloud account. Retool offers self-hosted deployment too, but only on its Enterprise plan, where its published page says Enterprise customers can deploy on their own infrastructure in less than 10 minutes. If self-hosting on a mid-tier budget matters, that is a real difference.
If that ownership model matches how your team works, talk to the Reflex team about a pilot build.
Best use cases for Reflex
A permit-and-parcel targeting engine
The core job is joining public permit filings and parcel data with your own service-area records and scoring every neighborhood. Reflex's Python backend handles the ingest, the join, and the scoring in one codebase, and the same framework renders the ranked-neighborhood dashboard. When the scoring model changes, it changes in git, with a test proving the new rules produce the expected ranking. How Retool compares: Retool can build the same dashboard quickly from SQL queries, but the scoring logic ends up distributed across queries and builder JavaScript rather than one testable module.
Write-back to the CRM and scheduling tools
Targeting is not read-only. A rep works a neighborhood, the outcome needs to land back in the CRM, and the next campaign needs to exclude that area. Reflex apps connect to your databases, APIs, and auth providers directly, and the Python backend executes whatever write-back logic you define. How Retool compares: Retool connects to REST APIs and databases well, and its Retool API publishes a rate limit of 300 points per 60-second window, which is worth checking against your write-back volume.
Self-hosted deployment for customer data
A contractor's customer list and job margins are commercially sensitive. Because the Reflex framework is free (Apache 2.0) for building and self-hosting apps, you can run the targeting app entirely inside your own cloud account and pay nothing to the vendor for the framework itself. How Retool compares: Retool supports self-hosting, but only on the Enterprise plan with custom pricing, so the self-hosted path is a sales conversation rather than a download.
AI-assisted first builds
Reflex's AI Builder plans an app before writing code, builds routes, state, and interfaces, and runs tests before handoff, which shortens the path from "here are our targeting rules" to a working draft. Company Context ingests your business rules and terminology so generated apps follow your conventions. How Retool compares: Retool's free plan includes 250 pooled AI credits per month and its Enterprise plan includes 10,000, so AI assistance is available there too, metered by credits.
Best use cases for Retool
Fast admin panels over stable data
If the targeting app is mostly a CRUD view, filter parcels, edit a campaign list, assign reps, Retool's drag-and-drop canvas over SQL is genuinely the faster path. Its published free plan covers up to 5 users, and the Team plan's published rate of $10 per builder per month is low for a small marketing-ops team. Reflex can build the same screens, but it is a framework, not a canvas, so the first version takes longer.
Teams without Python developers
Retool is built for operations and analyst teams who wire queries and components visually, with no codebase to own. If nobody on the team writes Python and nobody wants to maintain a repository, Retool's model fits better than Reflex's. Reflex's AI Builder narrows that gap, but a framework still assumes someone owns the code.
Platform overview of Reflex
Pure-Python full-stack apps
Reflex lets you build and deploy full-stack web apps, frontend, backend, and database, in pure Python. Components compile to JavaScript for the browser; non-UI logic lives in State. For a targeting app, that means the permit ingest, the scoring model, and the map interface are one codebase. Start with the getting-started introduction.
import reflex as rx
class TargetingState(rx.State):
min_permits: int = 3
zipcodes: list[str] = []
def neighborhood_card(z: str, score: float):
return rx.card(rx.text(f"{z}: score {score:.1f}"))
def index():
return rx.vstack(
rx.foreach(TargetingState.zipcodes, neighborhood_card)
)
Integrations with your data sources
Reflex integrates with databases, models, APIs, and authentication without a separate JavaScript integration layer. A targeting app reads permit APIs, parcel databases, and your CRM through the same Python backend. See reflex.dev for the integration overview.
AI Builder
Describe the app in plain language and the AI Builder plans it before writing code, builds routes, state, and interfaces, and runs tests before handoff. For a contractor, that means a working draft of the targeting dashboard from a rules document, not a blank repository.
Company Context
Feed in team documents, business rules, and design guidelines so generated apps follow your terminology and standards. A targeting app built with Company Context uses your service-area definitions and campaign vocabulary, not generic defaults.
Workflows with human review
Custom workflow apps are available today, and a dedicated Workflows product for AI agents with human review and approval is on the way. That fits campaign targeting directly: an agent drafts the neighborhood shortlist, a marketing manager approves it before spend is committed.
Flexible deployment and billing
Run on Reflex Cloud or your own infrastructure. Reflex Cloud billing combines organization seats and app compute, and compute charges depend on the resources your apps use and how long they run (see the billing docs). The framework itself is free (Apache 2.0) for building and self-hosting apps; the Reflex platform and AI Builder are sold as Enterprise with custom pricing, per Reflex Cloud.
How to choose decision matrix
| Your situation | Choose |
|---|---|
| Targeting rules change monthly and need tests and code review | Reflex |
| You need a CRUD panel over a stable data set this week | Retool |
| Your team writes Python but has no frontend specialists | Reflex |
| Your team is strictly non-technical analysts | Retool |
| You must self-host without an Enterprise contract | Reflex (Apache 2.0 framework) |
| You want published per-seat pricing and a free tier for 5 users | Retool |
| You need write-back to CRM and scheduling systems | Reflex |
| You want air-gapped deployment | Retool (Enterprise) or Reflex self-hosted |
| Budget is usage-based compute rather than per-seat | Reflex Cloud |
| Budget is predictable per-seat monthly cost | Retool |
Migration guide: switching from Retool to Reflex
Step 1: inventory the logic
Export every Retool query, transformer, and JavaScript block that touches targeting logic. Most targeting apps have three buckets: data connections (port directly), presentation (rebuild as Reflex components), and scoring logic (rebuild as Python functions). The third bucket is the reason to migrate; the first two are mechanical.
Step 2: set up the Reflex project
The framework is free under Apache 2.0, so setup is a pip install, not a procurement cycle.
pip install reflex reflex init
Step 3: port one query at a time
A Retool REST query calling a permit API becomes a Python function. Before, in Retool's builder, the query and its logic live in the canvas. After, in Reflex:
import os
import httpx
PERMIT_API_URL = os.environ["PERMIT_API_URL"]
async def fetch_permits(zipcode: str) -> list[dict]:
async with httpx.AsyncClient() as client:
resp = await client.get(
f"{PERMIT_API_URL}/records",
params={"zipcode": zipcode},
)
return resp.json()["records"]
The equivalent Retool call uses its REST resource with Bearer or OAuth 2.0 authentication configured in the platform. In Reflex, auth is your code: tokens live in environment variables your app reads.
Step 4: rebuild the interface
Rebuild the screens as Reflex components backed by State. The AI Builder can generate a first draft of the dashboard from a description of the Retool screens, then your team refines it.
What has no equivalent
Be honest about friction. Retool's visual builder, its prebuilt component library, and its published per-seat pricing have no Reflex equivalent; Reflex publishes no dollar rates, and its commercial platform is custom-priced Enterprise. Retool's air-gapped Enterprise deployment has no Reflex equivalent short of self-hosting the open-source framework yourself. If a team relies on Retool's builder for day-to-day edits by non-developers, a partial migration, keeping Retool for admin panels while moving the scoring engine to Reflex, is a legitimate end state.
FAQ
Is Reflex cheaper than Retool?
It depends on team size and hosting. Retool publishes a free plan for up to 5 users, $10 per builder per month on Team (annual billing), and $50 per builder per month on Business. Reflex's framework is free (Apache 2.0) for building and self-hosting; Reflex Cloud and AI Builder are custom-priced Enterprise with no published dollar rates. For a self-hosting team, Reflex's framework costs $0 in license fees; for a small cloud-only team, Retool's published rates are the lower published number.
Can a Reflex app pull public permit and parcel data directly?
Yes. Reflex integrates with databases, APIs, models, and auth providers from the Python backend, so the app can call permit APIs, read parcel databases, and join the results to your own service-area records in one codebase.
Does Retool support self-hosting?
Yes, on the Enterprise plan. Retool's published pages list cloud, self-hosted, and air-gapped deployment for Enterprise, and state that Enterprise customers can deploy on their own infrastructure in less than 10 minutes. Reflex supports self-hosting on the free open-source framework.
Do we need JavaScript to build a Reflex app?
No. Reflex apps are written in pure Python for frontend and backend; components compile to a React frontend and the backend runs on FastAPI. A Python-capable data or IT team can own the whole targeting app.
Which platform handles write-back to our CRM better?
Reflex, structurally. The Python backend executes whatever write-back logic you define against your CRM's API or database, with the logic versioned and tested. Retool also writes back through its REST resource (GET, POST, PUT, PATCH, DELETE), but the logic lives in the builder rather than your repository.
How long does a migration from Retool take?
Plan a pilot, not a big bang: port the scoring logic and one dashboard first, run both tools in parallel for one campaign cycle, and retire the Retool screens only when the team trusts the Reflex app. Data connections port quickly; interface rebuilds are the bulk of the work.
Campaign targeting is a competitive advantage, and the rules that drive it deserve code your team owns. Start a build with Reflex and put a working targeting app in front of your marketing team this quarter.