When a per-result scraper beats a Bright Data account
Last updated
Maintained by Datapika · We publish this product under the openclawai account on Apify. Product reference
The strongest Bright Data alternative for small and mid-size jobs is a per-result Apify actor: flat prices such as $0.005 per job listing or $2 per 1,000 Reddit posts, no subscription, no minimum order, and no company KYC. Bright Data's own entry points are a genuinely useful 5,000 free credits per month and datasets from $2.50 per 1,000 records with a $250 minimum, and its residential proxies require a verified company for zones created after July 7, 2026 (all checked September 6, 2026). Below is the honest breakdown, including when Bright Data is still the right call.
Why do teams look for a Bright Data alternative?
Three friction points recur. First, minimums and account structure: Bright Data's datasets carry a $250 minimum order, which buys 100,000 records at $2.50 per 1,000, and is the wrong shape when you need 2,000 rows for one analysis. Second, verification: since July 7, 2026, new residential proxy zones are available only to verified companies that pass a human-reviewed KYC check with a corporate email domain; personal Gmail or Outlook accounts are not eligible for residential access at all (docs.brightdata.com, September 6, 2026). That is a defensible anti-abuse policy, and also a hard wall for individual developers, researchers, and agent builders.
Third, product surface: Bright Data sells infrastructure, proxies, unblockers, scraper APIs, and datasets, and you assemble the solution. If what you actually want is 'give me the job listings with salary bands as JSON', an infrastructure account is a lot of machinery for one question.
None of this makes Bright Data bad; its free tier alone, 5,000 credits per month with no card, is more generous than most rivals. It makes it mis-shaped for per-question, per-run workloads.
- Dataset minimum: $250 per order at $2.50 per 1,000 records (brightdata.com, September 6, 2026)
- Residential proxies: company-only KYC for new zones since July 7, 2026; personal emails ineligible
- Infrastructure model: you assemble proxies, unblocking, parsing, and storage yourself
- Genuine strength: 5,000 free scraper-API credits per month, no credit card required
- The mismatch is shape, not quality: subscriptions and minimums vs one-off structured questions
How do per-result Apify actors compare to Bright Data?
Datapika's actors on Apify invert the model: instead of buying infrastructure and building a scraper, you call a finished vertical scraper and pay a flat price per delivered row. Job listings across Glassdoor, Indeed, LinkedIn, ZipRecruiter, and more cost $0.005 each with salary bands and cross-board dedup; Reddit posts cost $2 per 1,000 or $3 with full comment trees; TikTok, Douyin, and Bilibili videos cost $0.001 each (Apify billing records, September 6, 2026).
There is no subscription, no minimum, and no KYC: an Apify account with the free plan's $5 monthly credit is enough to pull about 1,000 job listings, and signup takes a personal email. A 20-row test costs $0.10, so the evaluation loop is minutes, not a sales call.
Structure is included rather than extra: rows arrive as typed JSON fields (salary_min, play_count, num_upvotes) instead of HTML or credits-per-request, so there is no parser to maintain when the target site changes its markup; that maintenance burden sits with the actor, and per-result billing means a broken actor earns nothing, which aligns incentives in your favor.
- Jobs: $0.005 per listing across multiple boards in one deduplicated run, salary bands included
- Reddit: $2 per 1,000 posts, $3 with comment trees; TikTok/Douyin/Bilibili: $0.001 per video
- Entry cost: $0; Apify's free plan carries $5 monthly credit, roughly 1,000 job rows
- No KYC, no corporate email requirement, no minimum order, no monthly commitment
- Typed JSON out of the box; no parser for you to maintain when markup changes
- Per-result billing: unsuccessful scrapes deliver no rows and earn the actor nothing
What about agents: MCP on Bright Data vs per-result actors?
Honesty first: Bright Data is not behind on agent tooling. Its Web MCP server is real, open, and has a free tier of 5,000 requests per month with no credit card, exposing generic tools for web search, page fetching with unblocking, and browser automation (brightdata.com, September 6, 2026). For an agent that needs 'read any page on the web', it is a strong offer.
The difference is generic versus vertical. Bright Data's MCP hands your agent a page-access capability; the agent still has to find, page through, and parse the target site, spending tokens and requests on navigation. A Datapika actor over the Apify MCP server (mcp.apify.com) hands the agent a finished tool with a typed input schema: 'jobs for this title in this city' or 'top posts in this subreddit' in one call, JSON rows back, billed per row.
For agent builders the billing shape matters as much as the tool shape: a flat per-row price is legible to an agent's cost budget (40 listings = $0.20, always), while credit systems that vary by rendering and target complexity are harder to bound in an autonomous loop.
- Bright Data Web MCP: generic web access tools, free tier of 5,000 requests/month, checked September 6, 2026
- Apify MCP + Datapika actors: vertical tools with typed schemas, one call per question
- Generic MCP spends agent turns on navigation and parsing; vertical tools return typed rows
- Flat per-row prices give agents a boundable cost per task ($0.005 x rows, no variables)
- Both routes need no credentials for the scraped sites; neither touches logins
How do Oxylabs and ScraperAPI fit into the comparison?
Oxylabs is Bright Data's closest peer: an infrastructure vendor whose Web Scraper API starts around $0.25 per 1,000 results on entry plans from $49 per month, bills only successful results, and offers an unlimited-time free trial of up to 2,000 results (oxylabs.io, September 6, 2026). Its billing became feature-based, with rates that vary by rendering requirements and target, which is efficient at scale and one more variable to model before you know a job's cost.
ScraperAPI is the budget generalist: $49 per month for 100,000 API credits on the Hobby plan, 1,000 free credits monthly, and a 7-day trial with 5,000 credits (September 6, 2026). Credits per request scale with rendering and premium features, and for most targets you get HTML back and parse it yourself; structured endpoints exist for a handful of big sites.
Both are subscriptions sized for steady monthly volume. Neither removes the parser-maintenance burden for arbitrary sites, and neither prices in a way an agent can bound per call as simply as a flat per-row actor. If your workload is a steady 500k+ generic requests per month, they get cheaper per unit than per-result actors; below that, the subscription idles.
- Oxylabs: from ~$0.25 per 1,000 results, plans from $49/month, pay only for successes, 2,000-result free trial
- Oxylabs billing varies by rendering and target complexity (feature-based, per its 2026 pricing update)
- ScraperAPI: $49/month for 100k credits, 1,000 free credits monthly; HTML out, your parser required
- Both are monthly subscriptions that idle below steady volume
- Per-unit cost beats per-result actors only once volume is consistently high
When should you stay on Bright Data?
Several cases are honestly better served by Bright Data. Sustained scale: at hundreds of thousands of pages per day, negotiated infrastructure pricing beats any flat per-row rate, and Bright Data's proxy network and unblocker are among the most capable available. Compliance-heavy procurement: its KYC process, enterprise agreements, and audit posture are features, not bugs, for banks, funds, and large enterprises that must show diligence on data sourcing.
Pre-collected datasets: when you need 10 million historical records from a major platform, a $250-minimum dataset purchase is efficient and immediate; no per-result actor will re-scrape that history cheaper. Arbitrary long-tail sites: per-result actors exist for specific verticals (jobs, Reddit, TikTok, Maps); if your target list is 300 assorted websites, an unblocking proxy or the Web MCP is the right primitive.
And frankly, tiny generic workloads: 5,000 free scraper-API credits per month, renewed monthly with no card, covers small recurring pulls at zero cost. The honest rule: choose Bright Data for infrastructure at scale or bulk history, choose per-result actors when the question is specific, structured, occasional, or asked by an agent with a budget.
- Stay for scale: 100k+ pages/day workloads where negotiated infra pricing wins
- Stay for compliance: KYC and enterprise agreements are assets in regulated procurement
- Stay for bulk history: multi-million-record dataset purchases have no per-result equivalent
- Stay for long-tail coverage: hundreds of assorted target sites need a generic unblocker
- Stay for free small pulls: 5,000 credits/month costs nothing and renews monthly
- Switch when the workload is specific, structured, per-question, or agent-driven
| Provider | Billing model | Entry cost | Free tier | Agent-readiness |
|---|---|---|---|---|
| Datapika actors (Apify) | Flat per delivered row ($0.005/job, $2/1k Reddit posts, $0.001/video) | $0, no minimum, no KYC, personal email fine | Apify free plan: $5 platform credit/month (~1,000 job rows) | Vertical MCP tools with typed schemas via mcp.apify.com; cost boundable per call |
| Bright Data | Credits per request (scraper APIs), $2.50/1k records for datasets, GB-based proxies | $250 minimum per dataset order; company KYC for new residential zones since July 7, 2026 | 5,000 scraper-API credits/month, no card | Web MCP with generic web tools, free 5,000 requests/month; agent parses results |
| Oxylabs | Per successful result, rate varies by rendering and target | Plans from $49/month | Free trial: up to 2,000 results, unlimited time | API-first; no flat per-row price, cost varies per call |
| ScraperAPI | API credits/month; credits per request scale with features | Hobby $49/month for 100k credits | 1,000 credits/month free; 7-day trial with 5,000 credits | Returns HTML for most targets; agent or you must parse |
Sources you can inspect
Product documentation and implementation references. These references describe supported options and billing, not independent proof of performance. Confirm current details before running; examples and estimates are not guarantees.
How to do it
- 1.Classify your workload: specific vertical questions (jobs, Reddit, TikTok, Maps) point to per-result actors; bulk history or 100+ arbitrary sites point to Bright Data-style infrastructure.
- 2.Price one concrete task on each candidate: for example, 1,000 job listings with salary fields is $5.00 flat on https://apify.com/openclawai/job-board-scraper, versus a $250-minimum dataset order or a subscription month elsewhere.
- 3.Run the free options head to head: Apify's $5 monthly credit vs Bright Data's 5,000 credits vs a ScraperAPI or Oxylabs trial, on the same query, and compare rows delivered, fields populated, and effort spent parsing.
- 4.Commit to the cheapest option that returned complete data with the least glue code, and re-run this comparison when volume grows tenfold; the right answer changes with scale.
Questions, answered
Is Bright Data's free tier better than paying per result?
For small generic workloads, often yes, and this page says so plainly: 5,000 scraper-API credits per month with no card (September 6, 2026) beats paying anyone, if a generic request gets you what you need. The free tier does not include structured vertical output like deduplicated multi-board job rows with salary bands, does not cover residential proxies, and resets monthly without rollover. Many teams sensibly use both: Bright Data's free credits for ad-hoc pages, per-result actors for structured verticals.
Does Bright Data really require KYC?
For residential proxies, yes, with a clear scope: new residential zones created after July 7, 2026 are available only to verified companies that pass a human-reviewed check, signing up with a corporate email domain; personal Gmail or Outlook accounts are not eligible, and pre-existing verified zones keep working (docs.brightdata.com, September 6, 2026). The scraper APIs and free tier do not carry that gate. Apify actors have no KYC at any tier; a personal email and the free plan are enough to run paid actors.
What is the cheapest Bright Data alternative for job listings specifically?
By listed price on September 6, 2026: Datapika's multi-board scraper at $0.005 per listing is $5.00 per 1,000 jobs with salary bands, dedup, and typed JSON, versus Bright Data's dataset route at $2.50 per 1,000 records but a $250 minimum order, which only wins if you genuinely need 100,000 records. For under roughly 50,000 fresh listings, per-result wins on total cost; for bulk historical archives, the dataset wins. Volume, not unit price, decides it.
Bright Data has an MCP server too; why use actors for agents?
It does, and its free 5,000 requests per month make it a fine generic web tool for agents (September 6, 2026). The difference is what the agent receives: Bright Data's MCP returns page content the agent must still navigate and parse across multiple calls, while a vertical actor tool returns finished typed rows, 40 job listings with salary fields for a flat $0.20, in one call. Fewer agent turns, boundable cost, no parsing logic in the prompt. Use the generic MCP for arbitrary pages, vertical tools for structured questions.
Are per-result actors reliable enough to replace an enterprise vendor?
Depends on the actor, so judge on published numbers rather than category. Datapika's job board scraper shows 28,721 runs across 2,595 users, and the Reddit scraper 7,185 runs with a 4.9 rating, per Apify records on September 6, 2026. What actors do not offer is enterprise SLAs, named support, or procurement paperwork; Bright Data and Oxylabs do. If your data feed is contractually critical, that support layer can be worth the premium; test both against the same queries first.
What extra costs apply on top of per-result actor prices?
Apify platform usage, billed by Apify: compute for run time, and proxy bandwidth where an actor defaults to residential proxies at $8 per GB, as the Reddit scraper does. Actor starts can carry a small event fee, typically $0.00005. For most runs these platform costs are cents against dollars of result fees, but they are not zero, so do what this site's own comparison pages recommend: run a bounded test, read the actual charge breakdown on your Apify usage page, then extrapolate.
Keep reading
Comparisons
Glassdoor API in 2026: real routes to jobs and salaries
Comparisons
Indeed API: What Exists, What Closed, What Works (2026)
Use cases
Web scraping API for AI agents (pay per result)
Use cases
Job Scraper API: Endpoints, Schema & Cost Math (2026)
Scrape guides
Reddit scraper without the API: $2 per 1,000 posts