Does Plus's Bigger REST Bucket Remove the GraphQL Migration Case?
Migrating to the GraphQL Admin API wins even on Plus, because a bigger bucket on a legacy API is still a legacy API. Plus multiplies the REST bucket tenfold, from 40 requests to 400 and from 2 per second to 20 (verified Sep 2026). Shopify has called REST legacy since October 1, 2024 and says all apps and integrations should be built with GraphQL. Treat the 10x as runway.
Your profile — see how the verdict shifts
- Confidence
- High — Read Shopify's REST Admin API rate-limit page on 2026-09-05 and confirmed the numbers exactly: bucket size 40 requests per app per store with a leak rate of 2 per second, and the bucket size and leak rate increased by a factor of 10 for Shopify Plus stores, giving 400 requests and 20 per second. The same page carries the sentence that decides this question: the REST Admin API is a legacy API as of October 1, 2024, and all apps and integrations should be built with the GraphQL Admin API. Our brief arrived with the opposite hypothesis, that the 10x bump makes migration optional. The page's own deprecation notice contradicts it, so we followed the page. Two further facts checked the same day sharpen it: Plus multiplies GraphQL throughput by the same factor, from 100 calculated cost points per second on standard plans to 1,000 on Plus, so the upgrade lifts both lanes rather than closing the gap between them; and bulk operations, the feature that actually solves high-volume reads, are a GraphQL Admin API capability. What the page does not state is a removal date for REST, which is exactly why runway is the honest word and panic is not. On the market side we checked the Store Management and Operations category and found no product sold against Admin API rate limits, because migrating REST code to GraphQL is engineering work rather than a purchase.
- Reference scenario
- $20M–$500M GMV · Shopify Plus · two or three REST-based integrations including an ERP connector polling orders · no REST removal date announced · in-house or agency dev bench
- As of
- September 2026
Decision at a Glance
| Your profile | Verdict | Why |
|---|---|---|
| REST code already scheduled for replacement within 12 months | WAIT | Migrating code you are about to delete is waste. No REST removal date is published, and the 400-request Plus bucket carries a doomed integration to its retirement date comfortably (verified Sep 2026). |
| Steady REST integrations comfortably inside the 400-request bucket | BUILD | No throttling pressure means you get to choose the calendar, which is the best position this decision ever offers. Migrate deliberately now rather than under a deadline Shopify sets later. |
| One integration exhausting the 400-request bucket at peak | BUILD | At 20 requests per second sustained, more retries only reshape the failure. GraphQL prices by query cost rather than request count, and bulk operations replace the polling loop entirely. |
| REST-based third-party apps you don't control | WAIT | You cannot migrate a vendor's code. Ask each vendor which Admin API their app uses and what their GraphQL plan is, then weight the answer when the renewal comes around. |
What Plus REST Rate-Limit Increase Actually Drives
| Outcome | Impact | How it works |
|---|---|---|
| Operational efficiency | High | Throttled integrations turn into retry queues and manual re-runs, and each incident consumes engineering hours that would otherwise have paid down the migration. |
| Revenue — indirect | High | Order, inventory and price syncs that throttle at peak publish late, so the storefront sells against stale numbers on exactly the days volume is highest. |
| Data & insight | Medium | GraphQL returns only the fields requested, which makes payload contracts explicit and downstream warehouse and reporting pipelines cheaper to maintain. |
| Customer experience | Medium | Shoppers meet this decision as oversells and delayed status updates when a throttled sync falls behind during a launch. |
Spend ceiling: Size the spend to the endpoint count, not the fear. Freezing new REST work costs nothing and stops the bill growing; a two-endpoint connector migrates inside the $10–25K contact-form band, and only a multi-system estate with polling rework justifies the top of the $15,000–$60,000 range (Deploi estimate, illustrative).
What buying enables (top apps)
- + Ten times the REST headroom on Plus, at 400 requests per app per store and 20 per second, with no code change at all (verified Sep 2026)
- + Existing integrations, vendor apps and internal tools keep working unchanged, since no removal date is published
- + Time: the bucket increase is real runway for a team that genuinely cannot fund a migration this year
- + Vendor-owned REST apps continue functioning regardless of what your own roadmap decides
What building additionally unlocks
- + Bulk operations for large reads, which move volume out of the request path instead of rationing it
- + Cost-priced queries that fetch related objects in one call, rather than one call per resource
- + Alignment with the API Shopify actively develops and documents new capabilities against
- + An undated platform risk removed from your register, which is the part no bucket size ever fixes
Find Your Verdict in 3 Questions
Is the REST code scheduled for replacement or retirement within the next 12 months?
Yes: Your verdict: WAIT — don't migrate code you're about to delete; the 400-request Plus bucket carries it to its retirement date.
No: Go to question 2.
Is the REST code yours to change, rather than a third-party app's?
Yes: Go to question 3.
No: Your verdict: WAIT — ask each vendor which Admin API their app uses and what their GraphQL plan is, then weight the answer at renewal.
Does any integration approach 400 requests in the bucket or 20 per second at peak?
Yes: Your verdict: BUILD — migrate that integration first and replace its polling loop with bulk operations ($15,000–$60,000, Deploi estimate, illustrative).
No: Your verdict: BUILD — you have the rare luxury of choosing the calendar; migrate deliberately now instead of under a deadline later.
The TCC Scorecard — 12 Dimensions
TCC — Total Cost of Capability: what it actually costs to have this capability over three years, whichever way you get it. Each dimension is scored 0–5 for both paths. How we score →
| Dimension | Buy | Build | Why |
|---|---|---|---|
| Cost | |||
| Acquisition & implementation | Staying on REST inside the Plus bucket costs nothing today; rewriting two or three integrations against the GraphQL Admin API runs 6–14 weeks depending on surface area (Deploi estimate, illustrative). | ||
| Recurring fees | Neither lane carries a licence, because there is nothing to buy here; the recurring cost on both sides is engineering attention rather than software. | ||
| Maintenance & upgrades | REST code accumulates workarounds for a shrinking surface, while GraphQL code tracks the API Shopify actively develops and documents its new capabilities against. | ||
| Switching & exit | Every additional month on REST adds endpoints to the eventual migration; a completed migration ends the question and leaves nothing to switch off later. | ||
| Risk | |||
| Vendor risk | There is no vendor in this decision except Shopify itself, and its published instruction is unambiguous: build with the GraphQL Admin API. | ||
| Security & compliance surface | Both lanes use the same authentication, scopes and protected-data rules, so the migration neither widens nor narrows what your integration can reach. | ||
| Platform-deprecation exposure | This is the whole decision. REST has been a legacy API since October 1, 2024 with no removal date published, which means the risk is real, undated and entirely on your side. | ||
| Value | |||
| Fit to requirement | REST fits a request-shaped world where one call returns one resource; GraphQL fits an integration that needs three related objects in a single, cost-priced query. | ||
| Time to market | Doing nothing ships instantly, which is exactly why teams keep choosing it; the migration takes a quarter and delivers no new feature on the day it lands. | ||
| Performance & scale | A 400-request bucket draining at 20 per second is a hard ceiling on call count, while bulk operations on the GraphQL Admin API move large reads out of the request path entirely. | ||
| Data ownership & AI-readiness | GraphQL returns exactly the fields you ask for, which makes payload contracts explicit and downstream data pipelines far easier to reason about than REST's fixed shapes. | ||
| Focus & opportunity cost | A migration wins no new customers and shows up on no roadmap, which is the honest cost of doing it before a deadline forces it. | ||
The App Landscape
| App | Status | Pricing | Best for |
|---|---|---|---|
| Shopify REST Admin API rate limits | Legacy — The REST Admin API is a legacy API as of October 1, 2024, and Shopify states that all apps and integrations should be built with the GraphQL Admin API. Rate limits are a leaky bucket per app per store: 40 requests with a leak rate of 2 per second on standard plans, increased by a factor of 10 for Shopify Plus stores to 400 requests and 20 per second (verified Sep 2026). No removal date is published on the page. | Included with every plan; the 10x multiplier is a Shopify Plus entitlement | Carrying existing REST integrations while you migrate them, not for anything new |
| Shopify GraphQL Admin API rate limits | Native — First-party Shopify, and the API Shopify tells developers to build on. Throughput is priced as calculated query cost rather than request count: 100 points per second on standard plans, 200 on Advanced, 1,000 on Shopify Plus, and 2,000 on Shopify for enterprise (Commerce Components) (verified Sep 2026). Plus multiplies this by the same factor of 10 it applies to REST. | Included with every plan; cost points per second are set by plan tier | Every integration you are building or rewriting from here |
| Workflow and integration tools (category) | Category — What the Store Management and Operations category actually contains, checked on 2026-09-05: general workflow, integration and reporting tools such as Zapier, Shopify Flow, Matrixify, Supermetrics and HubSpot. None is marketed as solving Admin API rate limits, and none can migrate your code, because they consume the same APIs your integrations do. Naming the absence is the finding: no app removes the REST migration. | Not applicable; no product in this category addresses Admin API rate limits | Moving specific workloads off custom code entirely, which sidesteps rather than solves the question |
| GraphQL migration (custom) | Build lane — The only lane that changes the answer. Inventory your REST calls, map each to its GraphQL equivalent, replace polling loops with bulk operations and webhooks, and retire endpoints in slices rather than in one release. Effort tracks the number of distinct endpoints, not the number of integrations. | $15,000–$60,000 one-time (Deploi estimate, illustrative), scaling with endpoint count | Any REST integration you own and expect to still be running in two years |
The Build Path
- Inventory the endpoints, not the integrations: Log every distinct REST path your systems call for a month, with call volume and peak rate against the 400-request bucket. Most teams discover that three or four endpoints account for nearly all traffic, which turns a frightening migration into a short, prioritized list.
- Replace polling with bulk operations and webhooks: The heaviest REST traffic is usually a loop asking whether anything changed. Bulk operations on the GraphQL Admin API move large reads out of the request path, and webhooks push changes instead of being polled for them. This usually cuts call volume before a single query is rewritten.
- Migrate in slices behind a shared client: Put both lanes behind one internal client so a single endpoint can move to GraphQL without a coordinated release. Slice by resource, run both paths against production for a week, and compare outputs before retiring the REST call.
- Effort band
- $15,000–$60,000 to migrate a typical mid-market REST estate to the GraphQL Admin API — Deploi estimate (illustrative); a two-endpoint ERP connector sits in the $10–25K contact-form band, while a multi-system estate with bulk-operation rework lands in $25–75K
- Typical timeline
- 6–14 weeks for two or three integrations, with the highest-volume endpoint moved in the first three (Deploi estimate, illustrative)
- Maintenance, honestly
- ~15% of build cost per year (Deploi estimate): roughly $2,500–$9,000/yr (Deploi estimate, illustrative), which is the same API version review a REST estate needs, against a surface Shopify is actively developing.
- What you own — and what you take on
- You own: the client, the query shapes, the cost budgeting and the schedule. You take on: a rewrite that ships no customer-visible feature, plus the discipline to finish it, since half-migrated estates carry both lanes' maintenance and neither lane's simplicity.
3-Year Total Cost of Capability
| Buy (app path) | Build (custom path) | |
|---|---|---|
| Year 0 (setup) | $0 (no action taken) | $15,000–$60,000 |
| Years 1–3 (recurring) | $9,000–$45,000 (incident response and REST workarounds) | $7,500–$27,000 (upkeep) |
| 3-year total | ≈$9,000–$45,000, with the migration still ahead of you | ≈$22,500–$87,000, with the migration behind you |
- † All figures illustrative samples for the reference scenario — not quotes, not verified pricing.
- † Buy column: staying on REST inside the Plus bucket, priced as the throttling incidents, retry engineering and workarounds a legacy estate accumulates. No software licence exists in this lane.
- † Build column: migrating two to three integrations to the GraphQL Admin API, including replacing one polling loop with bulk operations; three-year horizon.
What the Sticker Price Hides
On the buy path
- — The 10x Plus bucket reads like headroom and behaves like a deadline extension, since Shopify's legacy notice is dated October 1, 2024 and has no removal date attached
- — Plus lifts GraphQL by the same factor of 10, from 100 to 1,000 cost points per second, so the upgrade never makes REST relatively better (verified Sep 2026)
- — Every month on REST adds endpoints to the eventual migration, so the bill grows while the budget line stays invisible
- — Retry and backoff engineering built around a 400-request bucket is work you throw away the day the endpoint moves
On the build path
- — Half-migrated estates are the worst outcome: two clients, two error models and no simplification
- — Cost-point budgeting is a different discipline from counting requests, and teams that port REST call patterns straight across get throttled anyway
- — Vendor apps stay on whatever API their developer chose, so your migration never covers the whole store
- — ~$2,500–$9,000/yr upkeep, plus a rewrite that ships no visible feature (Deploi estimate, illustrative)
What Merchants Say
Teams upgrading to Plus describe expecting the API question to go away with the invoice, then finding the same integration throttled at a higher number during the same peak hour.
The recurring regret engineering leads report is timing rather than direction: everyone agreed GraphQL was right, and the migration only got funded once a peak-season incident made it urgent.
If You Change Your Mind Later
If you bought and outgrow it
Staying on REST has no exit, only a postponement. Nothing is stranded technically, since the endpoints keep working and no removal date is published, but the migration cost travels with you and grows as new REST calls are added. Freeze new REST work now, even if you migrate nothing this quarter, so the eventual bill stops compounding.
If you built and want out
Nothing strands, and the question closes. GraphQL client code, query shapes and cost budgeting are yours, they track the API Shopify actively develops, and a completed migration removes an undated platform risk from your register. The one thing to keep is the endpoint inventory, because it is the map that proves the migration is finished.
When This Answer Changes
We're watching for:
- ▸ Shopify publishing a removal date for the REST Admin API, which turns this from a calendar choice into a deadline
- ▸ A vendor app you depend on announcing its own GraphQL migration or failing to, which decides the part of your estate you don't control
- ▸ New Shopify capabilities shipping on the GraphQL Admin API only, which converts the migration case from throughput into feature access
Verdict change log:
- 2024-10-01Shopify declared the REST Admin API a legacy API and stated that all apps and integrations should be built with the GraphQL Admin API. That notice sits on the same page as the Plus rate-limit multiplier, which is why the bigger bucket cannot be read as an endorsement of staying. As of this page's September 2026 review, no removal date has been published, so the correct posture is a planned migration rather than an emergency one.
Common Questions
What are the REST Admin API rate limits on Shopify Plus?
Shopify Plus gets a bucket size of 400 requests per app per store with a leak rate of 20 per second (verified Sep 2026). Standard plans get 40 requests and 2 per second, and Shopify describes the Plus figures as the bucket size and leak rate increased by a factor of 10. The bucket is per app, so each app draws on its own.
Does the Plus rate-limit increase remove the case for migrating to GraphQL?
No. Plus multiplies the REST bucket tenfold and multiplies GraphQL throughput by the same factor, from 100 to 1,000 calculated cost points per second (verified Sep 2026). The gap between the two lanes is unchanged. Shopify has called REST a legacy API since October 1, 2024 and instructs developers to build on the GraphQL Admin API instead.
When will Shopify shut off the REST Admin API?
Shopify has published no removal date for the REST Admin API as of September 2026. The rate-limit page carries only the legacy designation dated October 1, 2024 and the instruction to build with the GraphQL Admin API. Plan a migration on your own calendar over the next few quarters, and stop adding new REST endpoints today, which costs nothing.
Your Next Steps
If you're going with BUILD(matches your selected profile)
- Log every distinct REST endpoint your systems call for a month, with call volume and peak requests per second
- Freeze new REST work today, so the migration stops growing while you plan it
- Put both API lanes behind one internal client so endpoints can move individually
- Move the highest-volume endpoint first, and replace polling with bulk operations and webhooks where it fits
- Run both paths in parallel for a week per slice and compare outputs before retiring the REST call
If you're going with WAIT
- Write down the retirement date for each REST integration you plan to replace rather than migrate
- Ask every app vendor which Admin API their app uses and what their GraphQL timeline is
- Add REST removal to your platform risk register with a quarterly review, since no date is published
- Re-decide the moment Shopify announces a removal date or a peak-season incident hits the bucket
Official Docs & Sources
- REST Admin API rate limits — shopify.dev
- Admin API rate limits (GraphQL) — shopify.dev
- Bulk operations with the GraphQL Admin API — shopify.dev
Official documentation linked for verification — our verdicts and estimates are our own.
Related Decisions
Should You Build or Buy a Historical Order Export Tool?
Shopify emails any order export past 50 orders and estimates 400,000 items at around 4 hours. The mechanics are identical on every plan, Plus included.
Are Bulk Operation Timeouts a Platform Ceiling or Your Bug?
A Shopify bulk query allows five total connections, two levels of nesting and 10 days to finish (verified Sep 2026). Plus raises none of the three.
How Many Bulk Operations Can Run at Once on Plus?
Each app gets up to five simultaneous bulk query operations per shop on API 2026-01 and higher (verified Sep 2026). Separate apps never share one slot.
Will the Storefront API Hold Up on Your Next Flash Sale?
Shopify says real-buyer requests face no fixed per-minute limit on the Storefront API. The ceiling is checkout creation, and it fails with a 200, not a 429.
Should You Build or Buy a PIM on Shopify?
PIM on Shopify is a scale decision: metafields cover most catalogs, PIM apps win at multi-channel breadth, custom pipelines at ERP-grade complexity.
Ready to plan the GraphQL migration before it's urgent?
We start with a month of call logs, rank your REST endpoints by volume and peak rate, and give you a migration order where the first slice removes most of the traffic.
Contact us todayVerdict scored for the reference scenario above. Estimates are not quotes; app pricing carries its verification date and gets re-verified quarterly. Full scoring anchors: see the TCC methodology.
Read how we score these decisions (the TCC Framework). No affiliate links, no paid placement — no app vendor pays to appear here.