Bing Search API Retired: 6 Replacements Compared (September 2026)

Microsoft retired the Bing Search API on 11 August 2025. Its own retirement notice is blunt: existing instances would be “decommissioned completely”, and the product “will no longer be available to be used or new customer signup”. There is no Bing search API on Microsoft's price list any more, at any price. What Microsoft sells in its place is a grounding tool for language models — priced, but not the same product. Checked 2026-09-16.

I read every rate below off the vendor's own live pricing page on 16 September 2026 and linked each one inline. I ran no benchmark and I make no claim about anyone's speed or success rate — this is a pricing and coverage comparison built from published rate cards and documentation, and the migration notes are estimates of developer effort, not measurements. Serpent is the API this site sells, and it appears here as one row among several.

What Microsoft retired, and what it sells instead

Microsoft's lifecycle announcement, dated 15 May 2025, states it in one sentence: “Bing Search APIs will be retired on August 11, 2025. Any existing instances of Bing Search APIs will be decommissioned completely, and the product will no longer be available to be used or new customer signup.” The same notice tells customers to migrate to Grounding with Bing Search as part of Azure AI Agents.

That is the whole of it. This is not a price rise or a tier shuffle — the product was withdrawn from sale. Teams that used Bing for rank tracking, content analysis, affiliate research or agent grounding all landed in the same place: find another supplier of Bing results, or stop using Bing results.

How I compared these replacements

The rules I held myself to, because this category is full of comparisons that quote one vendor's volume floor against another vendor's entry price:

• Entry rate and lowest published rate are separate columns, and the monthly commitment that unlocks the lowest rate is named beside it. A $1.96 rate that needs $106,050 a month is not cheaper than a $0.60 rate you can buy today.
• Only published rates. If a vendor never prints a per-1,000 number, I divide its own plan price by its own plan volume and say so — and if it prints nothing at all, the row stays empty rather than being filled with a guess.
• Grounding with Bing gets its own row and its own warning. It bills for grounded model output, not for a parsed results page, so putting it in the same column as a SERP API and letting you assume they are the same product would be the single most misleading thing this page could do.
• Exclusions are named in the table caption, with the reason. Quietly dropping a vendor that does not flatter the conclusion is selection bias.

The replacement table: published rates per 1,000

Arithmetic, checked 2026-09-16. SerpApi prints no per-1,000 rate at all, so two of its figures are our own division: $25 a month for 1,000 searches on Starter is $25.00 per 1,000, and $106,050 a month for 54,000,000 searches is $1.96 per 1,000. Bright Data's $1.30 is the rate it prints for requests beyond the 380,000 included in its $499-a-month plan; spread across the included volume that plan works out at $1.31 per 1,000. DataForSEO prints $0.0006 per SERP, which is $0.60 per 1,000 Bing SERPs. Brave: $5 of monthly credit at $5 per 1,000 requests is 1,000 requests a month. Serpent: $0.60 per 1,000 web search calls at the Default tier and $0.03 per 1,000 web search calls at the Scale tier, the same rate whichever engine you pick.

1. Serpent API — a drop-in engine parameter

Serpent exposes Bing through the same /api/search endpoint as every other engine, so the migration is a parameter and a field map rather than a new integration. Organic results, ads, related searches and sitelinks all come back in one JSON shape.

Migration, estimated: map webPages.value[] to results.organic[], rename a few fields, ship. Budget about two hours for a typical app.

# Before (Bing Search API, retired 11 August 2025)
import requests
r = requests.get(
    "https://api.bing.microsoft.com/v7.0/search",
    headers={"Ocp-Apim-Subscription-Key": KEY},
    params={"q": "best protein powder", "mkt": "en-US"}
)
results = r.json()["webPages"]["value"]

# After (Serpent API)
import requests
r = requests.get(
    "https://apiserpent.com/api/search",
    params={"q": "best protein powder", "engine": "bing",
            "country": "us", "api_key": KEY}
)
results = r.json()["results"]["organic"]

Cost: Bing bills from the web category, so the engine you pick never changes the rate: $0.60 per 1,000 web search calls at the Default tier, $0.06 per 1,000 web search calls at Growth, $0.03 per 1,000 web search calls at the Scale tier. Growth unlocks at a single $100 deposit and Scale at $500. Every new account gets free calls, shared across every free-eligible endpoint. Checked 2026-09-16.

Where it is weaker: five engines only — Google, Bing, Yahoo, DuckDuckGo and Brave. If your migration needs Yandex, Baidu, Naver or Seznam in the same SDK, three of the rows below cover more ground than we do. See the API reference for the full parameter list.

2. DataForSEO Bing Organic SERP API — $0.60 per 1,000, no plan to buy

DataForSEO ships a dedicated Bing endpoint rather than a Bing flag on a Google product, and it prices it the same way it prices Google: by queue. The Standard Queue is asynchronous — you post a task and collect the result later — while Live mode answers in the same request and costs more.

Cost: DataForSEO charges $0.60 per 1,000 Bing SERPs on its Standard Queue. The Priority Queue is $1.20 per 1,000 and Live mode is $2.00 per 1,000. One SERP is ten results, which the product page states in words. There is no volume ladder on this product, and DataForSEO's minimum payment is $50. Checked 2026-09-16.

Migration, estimated: rewriting a synchronous Bing call into the task-and-collect flow is the real work here, perhaps six hours; Live mode keeps you synchronous at over three times the Standard price. Our DataForSEO pricing explainer walks through the queue model in more detail.

3. SearchApi — stated per-1,000 rates on every plan

SearchApi is the one vendor on this page that prints a per-1,000 rate beside every plan instead of leaving you to divide, which makes it the easiest to budget against. Bing sits in its documented engine list alongside Google, Baidu, Naver, Yahoo, Yandex and DuckDuckGo.

Cost: SearchApi's entry rate is $4 per 1,000 searches on its $40-a-month plan, and its largest published plan is $5,000 a month for 5,000,000 searches, which it prints as $1 per 1,000 searches. New accounts get 100 free requests. Checked 2026-09-16.

Where it wins and loses: plan prices are monthly commitments, so a low rate is only available if you can fill the plan. Below about 10,000 searches a month you are paying $4 per 1,000, which is more than six times what DataForSEO charges for the same job.

4. Bright Data SERP API — seven engines, enterprise onboarding

Bright Data names seven engines on its SERP API product page — Google, Bing, DuckDuckGo, Yandex, Baidu, Yahoo and Naver — which is the widest coverage on this page. The trade-off is the way you buy it.

Cost: Bright Data charges $1.50 per 1,000 requests pay-as-you-go, with 5,000 requests a month on a free tier. Bright Data's $499-a-month plan includes 380,000 requests and prints $1.30 per 1,000 for requests beyond them; spread across the included volume that plan works out at $1.31 per 1,000. Checked 2026-09-16.

Migration, estimated: the API itself is straightforward. Know-your-customer checks and a sales conversation are part of onboarding, so plan in days rather than hours if you are starting from zero.

5. SerpApi — one ladder, no per-engine surcharge

SerpApi documents a Bing Search Engine Results API with its own engine=bing value, and the shape is mature. It publishes a single monthly ladder and no per-engine surcharge, so Bing costs what Google costs. What it does not publish is a per-1,000 rate.

import requests
r = requests.get("https://serpapi.com/search", params={
    "q": "best protein powder",
    "engine": "bing",
    "api_key": KEY,
})
results = r.json()["organic_results"]

Cost: SerpApi publishes one ladder for every engine and prints no per-1,000 rate at all, so these are our own divisions: $25 a month for 1,000 searches on Starter is $25.00 per 1,000; $75 a month for 5,000 searches on Developer is $15.00 per 1,000; and $106,050 a month for 54,000,000 searches, its largest published plan, is $1.96 per 1,000. Checked 2026-09-16.

Why teams still pick it: breadth. Its catalogue runs to more than a hundred named APIs, including Apple App Store, Walmart, Yelp and Google's verticals. If your migration needs those in one SDK, the price is the price. It also offers custom volume plans on request, with no published figure attached. Our SerpApi pricing explainer unpacks the ladder plan by plan.

6. Brave Search API — a different index, not a route to Bing results

Brave runs its own crawl and its own index. That makes it a genuine alternative source rather than a second route to Bing results, and it is the right answer if you were using Bing for independence from Google rather than for Bing specifically. It is the wrong answer if you needed Bing's SERP features to line up with Google's.

Cost: Brave: $5 per 1,000 requests on its Search plan. Every metered plan includes $5 in free credits every month, and $5 of credit at $5 per 1,000 requests is 1,000 requests a month — so the allowance did not disappear, the old named ladder did. Brave's own page today lists no plan called Free, Base or Pro; treat any older quote of that ladder as stale. Checked 2026-09-16.

If you are weighing an independent index against Google's, our Brave versus Google SERP data comparison and the overlap study put numbers on how far apart two engines actually sit.

Grounding with Bing Search: priced, but not a SERP API

This is the row most migration posts leave out, and it is the one Microsoft itself points at. Grounding with Bing Search lives inside Azure AI Foundry and is designed to be called by a language model, not by your rank tracker.

Cost: Grounding with Bing Search and Grounding with Bing Custom Search are both priced at $14 per 1,000 transactions, with a published capacity of 150 transactions per second and 1 million transactions per day. No free tier is shown on that page. Checked 2026-09-16.

What counts as a transaction is defined on the same page: “When the model reasoning determines it needs Grounding with Bing Search or Grounding with Bing Custom Search, the tool is invoked, and the result includes that data coming from Bing.” So you are billed per tool call the model decides to make, and what comes back is grounded model output with citations.

Put next to the SERP rows, the gap is large. That is roughly 23 times DataForSEO's $0.60 per 1,000 Bing SERPs, and nine times Bright Data's $1.50 per 1,000 requests. But the honest point is not that it is expensive — it is that it is a different product. If your app was an LLM app calling Bing for grounding, this is a real option and the price may be fine. If your app needed organic positions, snippets and SERP features, nothing in this section will give them to you, and the rows above are where you should be looking. Our guide to web search APIs for AI agents and RAG covers the grounding use case on its own terms.

The migration checklist

Whichever provider you pick, do these in order:

1. Diff the JSON shape. Run the same query through your old Bing call and the new provider, and map every field your app reads. Most schemas converge; the field names do not.
2. Verify SERP-feature parity. If you read People Also Ask, related searches or sitelinks from Bing, confirm the replacement returns them. Not all do, and a missing block is easy to miss when the response still looks valid.
3. Read the published throughput, not the marketing number. Brave publishes 50 queries per second on its Search plan and 2 on Answers; Grounding with Bing publishes 150 transactions per second and 1 million transactions per day. Where a vendor publishes nothing, assume nothing.
4. Check what happens to a failed request. Some vendors state plainly that they bill only for successful responses; others say nothing at all. Read the billing line on the vendor's own page before you budget, because the difference shows up in the invoice and not in the docs.
5. Watch the first week's spend. Per-call pricing behaves differently from a monthly plan at real volumes. Keep the old integration switched off but not deleted until an invoice has confirmed the model you expected.

Which option fits which job

• You need Bing organic positions and you are watching cost. The Serpent Bing endpoint, or DataForSEO's Standard Queue if you would rather buy from a specialist data vendor and can live with the task-and-collect flow.
• You want a stated per-1,000 rate you can put in a spreadsheet. SearchApi, which prints one beside every plan.
• You need Yandex, Baidu, Naver or Seznam alongside Bing. Bright Data or DataForSEO.
• You need Bing plus App Store, Walmart or Yelp in one SDK. SerpApi. It is the most expensive per search on this page and also the broadest.
• You used Bing for independence from Google, not for Bing. Brave Search API, and accept that the index is smaller.
• Your app is an LLM that just needs grounded answers. Grounding with Bing Search, at $14 per 1,000 transactions, is the supported path and Microsoft wrote it down as such.

FAQ

Did Microsoft really retire the Bing Search API?

Yes. Microsoft's own retirement notice, dated 15 May 2025, says Bing Search APIs were retired on August 11, 2025, that existing instances would be decommissioned completely, and that the product would no longer be available to be used or for new customer signup. Checked 2026-09-16. There is no Bing search API on Microsoft's price list today, at any price.

What replaced the Bing Search API?

Grounding with Bing Search, inside Azure AI Foundry, which Microsoft's own notice points customers at. Microsoft prices it at $14 per 1,000 transactions, and a transaction is counted when the model decides it needs the tool and the result includes data coming from Bing. It returns grounded model output with citations, not a parsed SERP, so it is not a drop-in replacement for code that read organic positions. Checked 2026-09-16.

What is the cheapest Bing Search API replacement in 2026?

On published rates, the Serpent web category is the lowest on this page at $0.03 per 1,000 web search calls at the Scale tier, which unlocks at a single $500 deposit. The cheapest third-party published rate here is DataForSEO at $0.60 per 1,000 Bing SERPs, with no monthly plan to buy and a $50 minimum payment. Checked 2026-09-16.

Can I migrate off the Bing Search API without rewriting my app?

Mostly. Most replacements return a similar JSON shape, and the main work is mapping field names: Bing's webPages.value becomes results.organic on Serpent and organic_results on several others. Budget two to four hours for a typical app, plus a pass over anything that read a SERP feature Bing had and your replacement may not.

Does Bing still return different results from Google?

Yes. Bing runs its own crawl and its own ranking, so its top ten is not Google's top ten. If you were using Bing because it surfaced pages Google did not, a Google-only replacement will not give you that back. Our search engine overlap study puts measured numbers on how far apart two engines sit.

Is there still time to migrate off the Bing Search API?

No. The deadline has passed: Microsoft's notice put the retirement at August 11, 2025, and says existing instances would be decommissioned completely. Checked 2026-09-16. Anything still pointed at the old endpoint is calling something Microsoft says is gone, so treat this as a migration you are already late for rather than one you are planning.

Related reading

If you are writing the migration now, the Bing Search API Python tutorial has the request-by-request version of the code above, and scraping Bing for free covers the honest limits of doing it yourself. For the wider price picture, see the SERP API pricing comparison, the walkthrough on migrating off SerpApi or DataForSEO, and the companion piece on Google Search API alternatives, which faces the same problem from the Google side.