Back to Learn

    Replit Alternatives: Six Tools Sorted by the Job You Used Replit For

    Replit is three products in one tab: an agent that builds apps from prose, a cloud IDE with a shell, and a host with a database. Most lists of alternatives mix replacements for all three, which is why they run to eight tools and no decision. This page sorts six alternatives by the job Replit was doing for you, prints each price as read on the vendor page on the date shown, quotes Replit's own pricing posts on why people leave, and shows what the Lovable route cost in our own measured runs.

    Updated

    Replit alternatives hero: the guide title beside a checklist card reading push the Repl to GitHub first, name the job, builder or IDE or pipeline, then re-run one prompt and measure it

    Why do people look for Replit alternatives?

    Two reasons are on the public record, and both are Replit's own words. The first is the meter. On 18 June 2025 Replit announced effort-based pricing for its agent: “Previously, the Agent charged $0.25 per checkpoint.” Under the new model, effort is “measured in terms of time and computation”, and “simple tasks may cost less than $0.25, more complex tasks may cost more than $0.25”. Replit's recap post of 12 July 2025 put numbers on both ends: “Simpler requests may cost as little as $.06” while “more involved requests will usually cost more than $0.25”, and disclosed a checkpoint billing error on 11 July that affected “approximately 6% of paying users”, refunded automatically with $10 in credits. The upper bound is not stated anywhere, and the cost is visible after the checkpoint, not before.

    The second is the incident. The most serious documented failure of any tool in this set. In July 2025 Replit's agent deleted a customer's production database during an active code freeze, generated fabricated data, and incorrectly reported that a rollback was impossible. Replit's CEO called it a catastrophic error of judgement. Fortune, 23 July 2025. The same section of our Lovable vs Replit page carries Lovable's own record, because no tool in this category is clean.

    The entry price, read on 26 September 2026 at replit.com/pricing: Core $20/mo ($18/mo billed annually), Pro $100/mo ($90/mo billed annually); no free plan is listed on the pricing page, which describes a Free Mode inside plans instead. Core includes “$20 towards powerful models” and Pro “$100 towards powerful models” per the plan cards; what a month of agent work costs past that is the effort meter above.

    Which job was Replit doing for you?

    Answer this before reading any list, ours included. Replit's own pricing page sells an agent (“10 parallel agents” on Pro), a workspace (“up to 60 projects”, “unlimited workspaces”) and hosting with a database (“28-day database rollback”) as one plan. Each has a different set of replacements, and a tool that is excellent at one is usually absent at the others.

    The three jobs Replit does and which of the six alternatives replaces each
     What you will missReplace it with
    Replit as an agent that builds the appProse in, working app with a database out, no shell neededLovable, Bolt.new, Base44
    Replit as a cloud IDE with a shellA terminal, a file tree, git, reachable from any machineCursor (desktop), GitHub Codespaces (browser)
    Replit as a step in a deploy pipelineAn agent that arrives in your review process rather than beside itv0 (pull requests into a Vercel project)

    One consequence is worth stating plainly. Nothing on this list does all three, which is the honest reading of why Replit costs what it does. Leaving it means picking the job you cared about and accepting a second tool, or a laptop, for the rest.

    The six, side by side

    Every price is quoted as printed on the vendor's page on 26 September 2026, linked in each card below. The existing-repository column is the one most lists skip and the one that decides whether leaving Replit is a move or a rebuild.

    Six Replit alternatives compared on category, metering, backend and whether they can open an existing repository; facts read 26 September 2026
     CategoryWhat gets meteredBackend built inOpens an existing repo
    LovablePrompt-to-app builder with a managed backendCredits, not seats, with unlimited workspace members on every planLovable Cloud ships database, authentication, storage, edge functions and realtime without a separate setup step, built on Supabase's foundationLovable's documentation supports exporting to GitHub but NOT importing an existing repository
    Bolt.newPrompt-to-app builderTokensYes, including databases and hostingGitHub integration is a separate feature from the ZIP export; the import path for an existing repository is not documented on the pages we checked, so we do not describe it
    Base44Prompt-to-app with a managed backendTwo meters: message credits for building and integration credits for runningYes, the most opinionated in this list: entities with CRUD, validation and security, serverless functions, configurable loginNot the direction it is built for
    CursorAI code editorIncluded monthly usage of Cursor's own models and of third-party models per plan; past the included usage you add on-demand use at API rates or move up a tierNone at allAny repository: clone it and open the folder
    GitHub CodespacesCloud development environmentCompute hours by machine size plus storageNone built inAny GitHub repository: it is the unit Codespaces attaches to
    v0Prompt-to-app inside the Vercel pipelineDollar-denominated credits spent at published per-model rates; generation pauses when they run out; unused credits expire after 65 daysYesYes, through pull requests against a repository you connect, which is the only review-first flow in this list

    Replit as an agent that builds the app

    Lovable

    Prompt-to-app builder with a managed backend

    You describe the app and it builds it, with Cloud (database, auth, storage, edge functions) provisioned as part of the build. The closest replacement for Replit Agent if you never opened the shell.

    Pricing
    Free tier plus paid plans priced in credits; we do not print Lovable's prices here because they change, and the pricing page is one click away, read on 26 September 2026. Source
    What gets metered
    Credits, not seats, with unlimited workspace members on every plan. Unused monthly credits roll over while the subscription is active, bounded by a two-month expiry. 5 build credits per day, up to 30 per calendar month, plus monthly Cloud and AI grants.
    Your existing Replit project
    Lovable's documentation supports exporting to GitHub but NOT importing an existing repository. If your app already exists, Lovable cannot open it.
    Backend, database and auth
    Lovable Cloud ships database, authentication, storage, edge functions and realtime without a separate setup step, built on Supabase's foundation.
    The documented weakness
    Two issues are on the public record: CVE-2025-48757 covering insufficient row-level security in generated projects, and an authorisation flaw in Lovable's own API disclosed in April 2026 that allowed a free account to read another user's source and database credentials. Reporting states it was patched for new projects but not remediated for projects created before November 2025.

    Bolt.new

    Prompt-to-app builder

    StackBlitz's builder. Prose in, a full-stack prototype out, with a ZIP export you run locally with npm.

    Pricing
    Free, Pro $25/mo, Teams $30/mo per member, read on 26 September 2026. Source
    What gets metered
    Tokens. Free gives 1M per month with a 300K daily cap; Pro starts at 10M per month with no daily cap. Paid tokens roll over for two months, free tokens do not.
    Your existing Replit project
    GitHub integration is a separate feature from the ZIP export; the import path for an existing repository is not documented on the pages we checked, so we do not describe it.
    Backend, database and auth
    Yes, including databases and hosting.
    The documented weakness
    Cost scales with codebase size rather than task size. Bolt's own documentation says tokens are consumed by reading, understanding and syncing your project files, so larger projects use more tokens per message. A one-line change to a big project is not a cheap request. Bolt token documentation

    Base44

    Prompt-to-app with a managed backend

    Chat and it builds on an opinionated managed backend with entities, functions and login built in. Owned by Wix since mid-2025.

    Pricing
    Free, Starter $16/mo, Builder $40/mo, Pro $80/mo, Elite $160/mo, all billed annually (monthly billing is 20% more), read on 26 September 2026. Source
    What gets metered
    Two meters: message credits for building and integration credits for running. The free tier gives 25 message and 100 integration credits per month across up to five apps.
    Your existing Replit project
    Not the direction it is built for. Its GitHub connection syncs a Base44 app out to a repository, and Base44's docs call that connection permanent and one-way.
    Backend, database and auth
    Yes, the most opinionated in this list: entities with CRUD, validation and security, serverless functions, configurable login.
    The documented weakness
    The GitHub connection is a one-way door in Base44's own words: it is “permanent” and “cannot be disconnected or transferred back to Base44”. Only a branch named main syncs. Base44 GitHub docs

    Replit as a cloud IDE with a shell

    Cursor

    AI code editor

    A desktop editor with an agent. The replacement if what you used in Replit was the shell, the file tree and an agent that edits a repository you own.

    Pricing
    Hobby free, Pro $20/mo, Pro Plus $60/mo, Ultra $200/mo, Teams Standard $40/user/mo, Teams Premium $120/user/mo, read on 26 September 2026. Source
    What gets metered
    Included monthly usage of Cursor's own models and of third-party models per plan; past the included usage you add on-demand use at API rates or move up a tier. Cursor's pricing docs no longer print the per-plan dollar allowances, so none are stated here.
    Your existing Replit project
    Any repository: clone it and open the folder. Nothing to import.
    Backend, database and auth
    None at all. No database, no authentication, no hosting, no deploy target. You bring the entire stack.
    The documented weakness
    The June 2025 Pro repricing replaced a request quota with a dollar allowance, and users hit limits within a few prompts on the stronger models. Cursor's CEO apologised that communication was not clear enough and offered refunds. TechCrunch, 7 July 2025

    GitHub Codespaces

    Cloud development environment

    VS Code in the browser on a container GitHub runs, attached to a repository. The replacement for Replit as a machine you can reach from any laptop, without an agent unless you add Copilot.

    Pricing
    GitHub Free includes 120 hours of compute and 15 GB-month of storage per month for personal accounts; beyond that, $0.18 per hour on a 2-core machine, $0.36 on 4 cores, $0.72 on 8, $1.44 on 16, $2.88 on 32, and $0.07 per GB-month of storage, read on 26 September 2026. Source
    What gets metered
    Compute hours by machine size plus storage. GitHub's docs: “the compute cost of using a codespace for an hour on a 16-core machine is eight times greater than a 2-core machine.”
    Your existing Replit project
    Any GitHub repository: it is the unit Codespaces attaches to.
    Backend, database and auth
    None built in. You bring a database and a host, exactly as you would on a laptop.
    The documented weakness
    No agent and no deploy target of its own. Everything Replit Agent did for you is now Copilot, a config file and a hosting decision. GitHub Codespaces billing docs

    Replit as a step in a deploy pipeline

    v0

    Prompt-to-app inside the Vercel pipeline

    Vercel's agent for Next.js, React, Tailwind and shadcn. It writes code and either deploys to Vercel or opens a pull request against your repository.

    Pricing
    Free with $5 credits, Plus $30/user/mo, Business $100/user/mo, read on 26 September 2026. Source
    What gets metered
    Dollar-denominated credits spent at published per-model rates; generation pauses when they run out; unused credits expire after 65 days.
    Your existing Replit project
    Yes, through pull requests against a repository you connect, which is the only review-first flow in this list.
    Backend, database and auth
    Yes. Vercel's docs are explicit that it builds backend logic, not just mockups.
    The documented weakness
    Seven messages per day on the free tier, and credits expire after 65 days. Vercel's docs state the $20 Premium tier is being sunsetted, so $30 per user is the entry paid price. v0 pricing docs

    What does the builder route cost, measured?

    Replit bills the agent in dollars per checkpoint; the builders bill in credits or tokens per message. The two do not convert, so the only honest number is one we measured ourselves. We ran three prompts from this site's library in a personal Lovable workspace and read the credits page before and after each. The prompts are the ones a Replit refugee would run first: a landing page with email capture, a set of empty and loading states, and a SaaS foundation with auth and roles.

    Credits one Lovable prompt consumed, three measured runs
    Landing page + email opt-in1.8 (one round)Empty states and skeletons9.5 (2 questions first)SaaS MVP with user auth10.4 (4 questions first)
    Credits one Lovable prompt consumed, three measured runs
    Categorycredits
    Landing page + email opt-in1.8 (one round)
    Empty states and skeletons9.5 (2 questions first)
    SaaS MVP with user auth10.4 (4 questions first)

    Read from the workspace credits page before and after each run; the two September runs asked clarifying questions first, which the prompts now invite. Data in src/data/runs.ts.

    Lovable's chat after the SaaS MVP prompt: the build-complete message listing sign-up, workspace invites and role checks, with the preview beside it
    The SaaS foundation run on 26 September 2026: Lovable asked four questions, enabled Cloud, then built and self-tested sign-up and invites. 10.4 credits, read from the credits page.

    The spread matters more than the average. A capture page was 1.8 credits in one round; a foundation with auth, roles and self-testing was 10.4. The prompt decides the bill, which is the same lesson Replit's effort meter teaches at a different price. The Lovable credits guide has the mechanics behind those numbers, and the prompt examples page shows what each run produced, screenshot by screenshot.

    What does the IDE route cost, and where is the free line?

    The IDE replacements meter time and machines rather than messages, which is easier to predict. GitHub's Codespaces billing page prints the whole table, and the free allowance on a personal GitHub Free account is 120 hours of compute and 15 GB-month of storage per month before any charge.

    GitHub Codespaces compute price per hour by machine size
    2-core machine0.184-core machine0.368-core machine0.7216-core machine1.4432-core machine2.88
    GitHub Codespaces compute price per hour by machine size
    CategoryUSD per hour
    2-core machine0.18
    4-core machine0.36
    8-core machine0.72
    16-core machine1.44
    32-core machine2.88

    Prices in US dollars per hour from GitHub's Codespaces billing documentation, read 26 September 2026. Storage is $0.07 per GB-month on top.

    Cursor is the other IDE route and it prices the agent, not the machine: Hobby free, Pro $20/mo, Pro Plus $60/mo, Ultra $200/mo, Teams Standard $40/user/mo, Teams Premium $120/user/mo, with included model usage per plan and on-demand use past it, per Cursor's pricing docs on 26 September 2026. If you go this way, the first thing to write is a rules file for the repository you just pushed; our Cursor rules guide has three you can copy.

    How do you move a project out of Replit?

    1. Push it to GitHub before anything else. Replit's Git pane documentation describes linking a Git provider and “Push updates: Send your commits to your remote repository with a single click”; the reverse direction, importing from GitHub, is documented too, so this step is reversible.
    2. Separate the code from the database. Replit's built-in database and secrets do not travel with the repository. Export the data and list the environment variables the app reads before you close the tab.
    3. Pick by the job, then by import. Cursor and Codespaces open the repository as it is. v0 works through pull requests against it. Lovable cannot import it at all, and Base44's GitHub link runs outward only, so those two are rebuilds from a prompt, which is a decision about how much of the Replit build was worth keeping.
    4. Re-run one representative prompt and read the meter. The free tiers listed in the FAQ are enough for one build each. What it costs on your own spec beats any figure on this page, including ours.

    What we deliberately did not publish

    • Replit's included credit amounts beyond the “$20 towards powerful models” and “$100” lines on the plan cards, and any average checkpoint cost, because the pricing page does not print them and third-party estimates disagree.
    • A same-prompt harness across all six tools. It needs paid accounts on each; without it nothing here ranks output quality, only mechanics and price.
    • CodeSandbox, which several lists include. Its pricing page refused our fetch on 26 September 2026, so no price could be verified and it is left out rather than quoted from memory.
    • Import mechanics for Bolt.new. The pages we checked document a ZIP export and a GitHub feature, not the path for opening an existing repository, so the card says so instead of guessing.

    Sources

    All checked on 26 September 2026. Shared vendor facts live in src/data/comparisons.ts and are dated on every page that uses them.

    Frequently asked

    What is the best Replit alternative?↓
    It depends on which of Replit's three jobs you were using. If Replit Agent built your apps from prose, Lovable or Bolt.new are the direct replacements, and Base44 if you want the most managed backend. If you used the shell and the file tree, Cursor on your laptop or GitHub Codespaces in the browser. If you want an agent that opens pull requests into a Vercel pipeline, v0. There is no single best; the lists that name one are selling it.
    Is there a free Replit alternative?↓
    Yes, several, and on 26 September 2026 Replit's own pricing page listed no free plan, only a Free Mode inside paid plans. Lovable's free tier gives daily build credits, Bolt.new has a free plan with a monthly token allowance, Cursor has a free Hobby plan, GitHub Codespaces includes 120 hours of compute a month on GitHub Free, and v0's free tier has $5 of credits with a seven-message daily cap. Each is enough to run one prompt and see the output before paying anything.
    Why is Replit expensive?↓
    Because of how the agent is metered. Since June 2025 Replit Agent bills per checkpoint by effort, which Replit defines as the time and computation used to fulfil each request; simpler requests can cost as little as $0.06 and more involved ones usually more than $0.25, with no ceiling stated. Replit's own recap post also documents a July 2025 billing error that affected about 6% of paying users and was refunded. You see the cost after the checkpoint, not before.
    Can I export my Replit project?↓
    Yes. Replit's Git pane connects a GitHub remote and pushes commits, and Replit's docs also describe importing from GitHub in the other direction. Push first, then decide where to open the repository: Cursor and Codespaces open any repository; Lovable cannot import an existing one, per its own docs, so a Replit project going to Lovable is a rebuild from prompts, not a move.
    Is Lovable better than Replit?↓
    For someone who never used the shell, usually. For someone who did, no: Lovable has no terminal, edits one branch at a time and cannot import a repository. The full comparison, with both tools' metering and their documented failures, is on our Lovable vs Replit page.
    Which alternatives did you actually run?↓
    Lovable. The credit figures on this page come from our own runs in a personal Lovable workspace, read from the credits page before and after, with screenshots. We did not run the same prompt in Bolt, Base44, Cursor, Codespaces or v0, because a fair harness needs paid accounts on all of them, so nothing on this page ranks output quality. Prices and mechanics are what the vendor pages say on the date printed.

    Related reading

    Written by

    Marco Kohns

    Marco Kohns

    Founder of ProtoBites - Venture Growth Studio

    Growth PM at a Silicon Valley scale-up (a16z and General Catalyst backed), ex-Techstars where he consulted 13 early-stage startups, Reforge-trained. Every prompt on this site comes out of shipping ProtoBites' own portfolio products.