6 Flatfile Competitors Worth Evaluating in 2026 (Flatfile Is Now Obvious)

Flatfile built the category. Embedded CSV and Excel import for SaaS onboarding — Spaces, Workbooks, Blueprints, listeners and jobs for validation and transformation, AI column mapping, and a plugin library covering everything from dedupe to PDF extraction. If you've shipped a "import your customers" button in the last five years, you've at least evaluated it.

Before comparing Flatfile competitors, though, there's a fact that changes the frame: Flatfile has renamed itself to Obvious.

The rename, and why it matters to your decision

Flatfile.com carries a site-wide banner saying the company is now Obvious, linking to obvious.ai. Archived snapshots place the change between early December 2025 and mid-January 2026. There was no press release — the most recent item in Flatfile's own news feed predates it.

Obvious is not a data importer. It's a general AI agent workspace with Boards, Sheets, Slides, Docs, Apps and meeting transcription, and its comparison content positions it against Notion, ChatGPT and Gemini. Different category entirely.

The Flatfile product line still exists and its docs remain excellent. But three things are worth putting side by side:

• "Portal" and "Embedded Data Import" have vanished from the site navigation. Where they used to sit, you now find Workbooks, Mapping, Transform, and solutions pages aimed at service teams, system integrators, NetSuite and Workday. The company is pointing at enterprise migration and partner channels, not at SaaS developers adding a CSV button.
• The front-end SDKs have gone quiet. As of August 2026, the JavaScript, listener and record-hook packages on npm last shipped somewhere between roughly nine and sixteen months ago. For comparison, CSVbox and Dromo both published releases within the last two weeks.
• Obvious publishes credit-based pricing. Flatfile does not publish any. The new product will tell you what it costs; the data-import product asks you to talk to sales.

What Flatfile actually charges

Nothing published. The pricing page lists feature bundles and two buttons: talk to sales, book a demo. The stated model is that pricing is based on project volume and explicitly not per seat, with development-environment usage free and billing on a monthly cycle. Self-managed and multi-region deployments are listed as enterprise features — available, but sales-gated with no self-serve path.

That opacity is the single most durable complaint in the public record. It goes back years: on the Hacker News thread launching CSVbox in 2021, one commenter's assessment was that Flatfile's pricing was out of whack for what the product did, and that being a cheaper version looked like a good market. Five years later, half this shortlist exists because that was correct.

The other durable complaint is data custody. The author of HelloCSV put it plainly in a 2025 thread: they liked Flatfile, but it sends data to remote servers, and as a healthcare company they needed full control over data storage. That single objection explains most of what follows.

The shortlist at a glance

Two organising axes run through that table, and they're more useful than any feature checklist.

The first is published prices versus talk to sales. CSVbox, Impler and UseCSV publish everything; Dromo publishes its main tier; OneSchema, Ingestro and Flatfile publish nothing. If procurement friction is what sent you looking, you've already narrowed the field.

The second is newer and more interesting: "self-hosting" has quietly been replaced by "the vendor never sees the data." Both CSVbox and Dromo now offer a private mode where the importer runs entirely in your end user's browser and posts straight to your endpoint. That answers the healthcare and GDPR objection without anyone running servers — which is a genuinely better outcome than the on-prem deployments people were asking for in 2021.

1. OneSchema — closest feature-for-feature competitor

OneSchema is the most direct like-for-like: an embeddable importer plus a newer recurring-file-pipeline product with AI agents, sold into healthcare, HR and benefits, manufacturing and system integrators. Region-specific hosting, business associate agreements, headless importer API, white-labelling, localisation, enterprise SSO and role-based access — the enterprise checklist is complete.

The problem is that it reproduces the exact friction you're leaving. Every tier says "contact us." Metering is by file consumption tiers rather than rows. White-labelling, localisation and EU hosting are gated to the middle tier rather than the entry one, and self-hosting is a paid enterprise add-on. Its product attention is also visibly shifting toward the agentic pipeline product — the same drift pattern worth noticing at Flatfile.

2. CSVbox — best value, and it's not close

CSVbox is what happens when someone reads the 2021 Hacker News thread and builds the cheap version. As of August 2026 its paid tiers run from roughly the price of a lunch to a few hundred a month depending on rows per import, all published, all self-serve, with a free sandbox for testing.

The parts that matter beyond price: schema and validation are configured in the admin UI rather than in code, so a non-engineer can adjust field rules without a deploy. SOC 2 Type II and GDPR posture apply on every plan, including the free one — many vendors gate compliance behind enterprise. And private mode means the importer can run in your user's browser and post directly to your endpoint, with rows never touching CSVbox servers. Its SDK packages are also the most actively maintained on this list.

The honest cons are about scale and organisation size. There's no on-premise deployment product despite it being asked for since 2021, team member caps are low, there's no enterprise SSO or SLA story, and it's a small vendor — which is a real continuity consideration for a core flow.

3. Dromo — best if the data is regulated

Dromo's whole pitch is data custody, and unlike most such pitches it's structural rather than promissory. Private mode processes imports entirely in the end user's browser, and the AI column matching happens there too — no shipping data to Dromo or to a third-party model provider. SOC 2 Type II, HIPAA with a business associate agreement, and GDPR apply across all plans, not just the top one, and white-labelling is included rather than upsold. True on-premise deployment is available at the enterprise tier.

The catch is the price floor. Dromo's published professional tier starts at several hundred dollars a month with per-import overage above its included allowance, plus caps on rows per import and number of schemas. For a startup adding a CSV button that's a lot; for a healthcare SaaS whose security review would otherwise kill the deal, it's cheap.

One credibility note: Dromo's public comparison pages carry a footnote saying the competitor data was collected in mid-2023. Don't take its claims about Flatfile's contract terms at face value — Flatfile's own FAQ describes monthly billing, which contradicts them.

4. Impler — the credible open-source option

Impler is MIT-licensed, actively maintained, and priced by rows rather than files — with a free tier and paid plans that undercut everything else hosted here. React, Angular and JavaScript SDKs, custom code validations, dynamic runtime schemas, webhook auth with retries, and scheduled auto-imports.

The MIT licence is the real differentiator: self-hosting is genuinely available, not an enterprise upsell. That's the answer for teams whose compliance requirement is "the data doesn't leave our infrastructure" rather than "the vendor has a certificate."

Which is the other side of it. Impler publishes no SOC 2, ISO or HIPAA posture that we could find, and it's a small project with a modest release cadence — meaningful key-person risk for something sitting in your onboarding funnel. If you self-host, the ops burden is yours.

5. Ingestro — the EU option, formerly nuvo

If you're working from an older comparison, note that nuvo has renamed to Ingestro. The old nuvo.codes domain no longer resolves at all, and getnuvo.com redirects. Any link you have is dead.

The product is an embeddable importer plus recurring data pipelines, with AI mapping including multi-language column matching and a cleaning assistant. Its strongest card is the compliance story: ISO 27001 certification and a GDPR-first posture aimed squarely at European buyers, which is a better fit than a US vendor's DPA for some procurement teams.

Two reservations. Pricing is not public — a base rate plus usage fees, quoted on a call — so it reproduces the problem you're solving. And the rebrand itself is a mild risk signal for a core dependency; it's the second company in this article to change its name inside a year.

6. Build it yourself — stronger than it was

The DIY lane got materially better. Papa Parse remains the parsing workhorse, MIT-licensed and actively maintained. HelloCSV was written specifically as an open-source Flatfile alternative by an engineer at a healthcare company, runs entirely in the browser, and is under active development. react-spreadsheet-import covers upload, parse, auto column matching, validation hooks and theming.

The trap is well documented and worth stating clearly: parsing is about five percent of the job. The permanent maintenance tax is Windows-1252 encodings, European date formats, embedded newlines inside quoted fields, multi-sheet workbooks, very large uploads, and server-side validation. And you have no SOC 2 artefact to hand a customer's security reviewer, which in B2B SaaS can cost you more in deal friction than the subscription you avoided.

When Flatfile is still the right call

Stay on Flatfile if you're already integrated, your volumes fit your contract, and your use case is enterprise data migration — because that's precisely where the company is now aiming, and the Workbooks, Mapping and Transform surfaces are the most sophisticated in this category by a distance. Nothing here matches its plugin library or its Blueprint model.

Reconsider if any of these are true: you're a self-serve SaaS whose importer is a small part of onboarding and you're paying enterprise rates for it; your customers' security reviews keep stalling on data residency; or you need SDK updates in a framework version that hasn't seen a release in a year. Those are concrete. "The company changed its name" is a prompt to look, not a reason to migrate on its own.

The verdict

Getting the data in is usually the first of several problems — what happens to it afterward is the next one. Our Improvado alternatives piece covers the reporting and pipeline layer, and Zapier vs Make is the practical starting point if your post-import routing is still glue code someone wrote in a hurry.