Most searches for an alternative start with a specific irritation rather than a general dissatisfaction. A base hit its record cap in the middle of an application window. The bill grew because two people from another department needed to see the list. A form view collected everything asked of it and there was still no way to reply to anyone from inside the tool.
Those three complaints point in different directions, and the comparison articles treat them as one. They are almost all ranked lists of eight products that do the same job, which is fine if the job is right. The first question worth answering is whether the replacement should be another database at all.
Name the thing that is actually wrong
Three different problems send people looking, and the right alternative depends entirely on which one is in play.
Volume. A base filled up. Airtable lists 1,000 records per base on the free plan, 50,000 on Team, 125,000 on Business, and 500,000 on Enterprise Scale. If the count is the whole problem, the answer is a product with a higher cap or one you host yourself, and nothing else about the setup needs to change.
Seats. The price scales with people, so the second reviewer, the finance colleague who needs to look, and the manager who reads the list on Fridays each add to the bill. Airtable lists Team at $20 per seat per month billed annually, or $24 billed monthly, and Business at $45 per seat per month billed annually, or $54 monthly. The free plan allows up to 5 editors in a single workspace, with 50 commenters. If seats are the problem, look at products that price a whole workspace rather than each person, or at a tool whose plan tiers include the people you actually have.
Shape. The tool stores the request and has no view of what happens to it. A form view creates a row, a single select column holds the status, and the reply is composed in a mail client where nobody else can see it. If this is the problem, a different database will not fix it, because it is a property of the category rather than of the product.
Sort the complaint into one of those three before reading a single comparison. Teams that skip this step migrate to a near identical product, spend a fortnight rebuilding the bases, and discover the same gap on the far side.
What the numbers actually are
Here are the published figures from each product's own pricing page at the time of writing, on annual billing where the page advertises a discount for it. Record caps are quoted in the unit each product uses, which is not the same unit across products and is the detail most comparisons flatten.
| Product | Free tier | First paid tier | Cap on that tier | Higher tier | Self host |
|---|---|---|---|---|---|
| Airtable | 1,000 records per base, 5 editors, 1 GB per base | Team, $20 per seat per month | 50,000 records per base, 20 GB per base | Business, $45 per seat per month, 125,000 per base | No |
| Baserow | 3,000 rows per workspace, 2 GB per workspace | Premium, $10 per user per month | 50,000 rows per workspace, 20 GB | Advanced, $18 per user per month, 250,000 rows | Yes |
| NocoDB | 1,000 records, 3 editor seats, 1 GB | Plus, $12 per seat per month | 50,000 records, 20 GB | Business, $24 per seat per month, 300,000 records | Yes |
| Grist | 5,000 records per document, unlimited documents | Pro, $8 per user per month | 100,000 records per document | Business, $24 per user per month, minimum 5 users | Yes |
Two things in that table deserve a second look.
Per base is not per workspace. A cap of 50,000 records per base, as Airtable's Team plan lists, is a much larger allowance than 50,000 rows per workspace once there are six tables in play, because the base cap resets with each new base. Read the unit before comparing the numbers.
Unlimited seats options exist. NocoDB lists its Plus tier at $12 per seat per month billed annually, and also offers unlimited seats at $108 a month, with the same choice on Business at $216. For a team of more than nine or ten, that changes the arithmetic completely, and it is the kind of line that a ranked list of eight products never quotes.
Revision history is the other figure worth checking against your obligations rather than your preferences. Airtable lists a two week revision history on the free plan, one year on Team, and two years on Business. Baserow lists 14 days of row change history on free, 90 on Premium, and 180 on Advanced. Grist lists 30 days of snapshots on free and three years on Pro. If a decision has to be defensible a year after it was made, a fortnight is not a history.
The question the comparison lists do not ask
Every product in the table above will collect a request through a form, show it as a row, let you set a status, and notify a chat channel. None of them will answer the person who sent it.
That sounds like a small omission and it is where most of the working day goes. Taking in a request has three parts: it arrives, somebody deals with it, somebody replies. A database is built for the first part and offers a column for the second. The third part happens in email, which means the record of what was said lives in one person's sent folder, next to their personal mail, invisible to everybody else. Six weeks later, working out what an applicant was told requires the cooperation of whoever told them.
The consequences compound. Nobody can see whether a reply went out, so the same person gets answered twice or not at all. Nothing records when the status changed, so the question of how long the average applicant waited cannot be answered from the data. Whether the message was opened is not knowable at all. And the same person appearing in three forms is three unrelated rows, because the key that would join them was never created.
A form tool with response management is built from the other end. The submission arrives with an owner and a stage attached, the send box sits on the same screen as the answers, every send stays on record with whether it was opened, and contacts build themselves from the email address so one person answering several forms stays one person. The trade is real and worth stating: this category gives up the modelling range of a database. If the work involves products, orders, and suppliers that have to reconcile against each other, the database is the correct tool and no amount of response management replaces it.
The pricing model differs too, and it differs in a way that suits a different situation. Where a database meters records, a tool in this category meters people: forms and responses unlimited on every tier, with published plans at $0 for one person, $39 a month for up to three, $99 for up to ten, and $249 for up to thirty, and two months free when paid yearly. A team of three collecting forty thousand submissions a year is expensive to serve on a record cap and cheap on a seat count. A team of twenty collecting four hundred is the reverse. Neither model is more honest than the other, but one of them matches the shape of your work and the other does not.
Four things to check before committing
These decide whether a replacement survives contact with a real queue, and none of them are a tick in a grid.
Whether a helper can be read only. Somebody outside the core team always needs to see the list without being able to change it. Airtable lists unlimited commenters on paid plans and 50 on the free tier. Baserow and NocoDB place role based permissions on their higher plans. Check what the cheapest workable arrangement is, because the alternative is giving a temporary helper full edit rights.
Whether uploaded files are protected by more than obscurity. Applications arrive with a CV, a certificate, a photograph of an identity document. The published attachment allowances are easy to compare. The question worth more is whether knowing the file's address is enough to open it, or whether the reader must be signed in first. Ask it explicitly, because it is rarely on the pricing page.
Where the data can go on exit. Every product in the table exports CSV, and CSV loses the relationships between tables. Before migrating in, find out what comes out: whether attachments are included, whether the links between tables survive, and whether there is an API that can be used to take everything rather than one table at a time.
How long the rebuild takes. Importing an Airtable base into another database is a supported path in most of these products and it is not free. Formulas use different syntax, automations have to be recreated, and any integration pointing at the old API needs redirecting. Budget the days, and count them against the annual saving before deciding the move is worth it.
What to compare, in order
Decide in this order and the shortlist writes itself.
- Category first. One kind of record arriving repeatedly, where the work is the reply, points at a form tool with response management. Several kinds of record that reference each other point at a database.
- Cost driver second. If the team is small and the volume is large, prefer a per seat or per workspace price over a record cap. If the team is large and the volume is small, prefer the record cap.
- Hosting third. If the data cannot leave your own infrastructure, the shortlist is the products that publish a self hosted edition, and the comparison is against the cost of the hours rather than against zero.
- Permissions and history fourth. Read only access and the length of the revision window, checked against what you are obliged to be able to prove.
- Exit fifth. What comes out, in what format, including attachments.
Feature grids belong at the bottom of that list, not the top. By the time the first four questions are answered, most of the eight product shortlists have collapsed to two.
What to change first
Work out which of the three complaints is yours before reading another comparison. If it is volume or seats, pick from the table above on the cap unit and the hosting model. If the real problem is that the tool has nothing to say about the reply, the features that matter are owner, stage, history, and a send box on the same screen, and one week of real requests through Halict will show whether that closes the gap.
Q1. Is there a genuinely free alternative with a higher record limit?
Grist lists 5,000 records per document on its free tier with unlimited documents, and Baserow lists 3,000 rows per workspace, against Airtable's 1,000 records per base. Self hosted editions of Baserow, NocoDB, and Grist remove the subscription entirely and add a server, backups, and upgrades in its place. Compare against the cost of those hours rather than against zero.
Q2. Can an existing Airtable base be moved without rebuilding it?
Most of these products publish an Airtable import, and it carries the tables and the data rather than the whole setup. Expect to recreate formulas in a different syntax, rebuild automations, and repoint anything that calls the old API. The import is the easy half, so budget for the second half before committing to a date.
Q3. Does a form tool with response management replace a database entirely?
Only for work that is one kind of record arriving repeatedly. It holds the owner, the stage, the history, and the sent reply, which a database does not, and it does not model several kinds of record that reconcile against each other. Teams running both are common: the queue in the form tool, the reference data in the database.
Q4. What is the cheapest way to let a colleague view the list without paying for a seat?
Check the commenter or view only role on each product before assuming a full seat is needed. Airtable lists 50 commenters on the free tier and unlimited on paid plans. Tools priced by plan tier rather than by individual seat often include a view only role at the level a team is already paying for, which makes the answer nothing extra.
Q5. Does hitting a record cap mean losing data?
Generally not, and the exact behaviour is in each product's terms, which is worth reading before relying on a tier. The usual pattern is that new rows stop being accepted while existing data stays readable. That is a soft failure arriving at the worst possible moment, in the middle of an intake window, so estimate the annual volume, double it, and compare that against the cap rather than against today's count.