Context.dev Review (2026): The Flat-Rate Scraping API, vs Apify and Firecrawl

Direct answer

Context.dev is a web-scraping and data-enrichment API for AI agents with unusually simple billing: 1 credit = 1 scrape, with JS rendering, anti-bot bypass and residential proxies included at no surcharge (verified on its live pricing page, July 2026). Plans run from a real free tier (500 credits a month with a work email) through Developer at $25/month for 10,000 credits to Scale at $499/month for 1 million. It bundles brand data, entity enrichment, page-change monitors and a hosted MCP server. The honest caveats: it's a solo-founder YC Summer 2026 company, there's no prebuilt-scraper marketplace like Apify's and no browser automation like Firecrawl's, and the independent track record is still thin — we have not run it on our own bench.

Updated

Jul 13, 2026

Evidence

3 checks

Sources

9 source links

Target query

context.dev review

Evidence used

• Every plan price, credit cost and rate limit was read directly from context.dev/pricing on Jul 13, 2026 — including the no-surcharge wording quoted in the pricing section.
• Company background verified against Y Combinator's company profile, the founder's Launch HN thread (Jul 9, 2026) and the 2024 Show HN for Brand.dev, its previous incarnation.
• Comparison figures come from Apify's and Firecrawl's live pricing pages, checked the same day.

How we checked this

• This is a research-based review: we have not yet run Context.dev on our own scraping bench, and nothing here claims first-party test results.
• All dollar figures and credit costs were read from the vendors' live pricing pages on Jul 13, 2026; the per-1,000-scrape comparisons are our arithmetic from those published numbers, and we label them as such.
• Where Context.dev's own docs disagreed with its pricing page (legacy tier names and rate limits linger in the docs), we used the pricing page, which matches current checkout.

What Context.dev actually is

Context.dev is one API that turns any URL into agent-ready data: scraping to markdown, web search, brand intelligence, entity enrichment and change monitors, with hosted and local MCP servers and SDKs in five languages.

The product splits into three API families plus a watcher. The Web API scrapes any URL to LLM-ready markdown or HTML, crawls sites, runs web search, pulls sitemaps and screenshots, and parses PDFs, Word and Excel files. Brand Intelligence returns a company's logos, colors, fonts and social handles from a domain, email or ticker. Entity Enrichment does the odder, stickier jobs: product extraction, NAICS/SIC industry classification, and mapping cryptic bank-transaction descriptors to real merchants. Monitors watch a page, sitemap or extraction for changes and fire webhooks. SDKs ship for TypeScript, Python, Ruby, Go and PHP, with Zapier, Make, Google Sheets and Excel connectors for the no-code side.

For AI-agent builders the headline is the MCP server, offered both hosted (a remote endpoint you point Claude, ChatGPT or Cursor at with an API key) and local via npx. It uses a code-mode design — the agent gets a docs-search tool and an execute tool that runs TypeScript against the SDK in a sandbox — rather than one MCP tool per endpoint. That's the same 'agents can use it directly' promise Apify makes for its hosted MCP server, applied to generic scraping instead of a scraper marketplace.

The company is younger than it looks in one direction and older in the other. The unified scraping product is months old, but it grew out of Brand.dev, a brand-data API that did its Show HN in mid-2024 — so there are roughly two years of production lineage behind the brand endpoints. It's a Y Combinator Summer 2026 company built by a solo founder, Yahia Bakour (ex-Amazon; co-founded and sold StockAlarm.io), whose stated ambition is "the context layer for the internet: one API that helps agents and software understand websites, companies, products, and brands." The July 2, 2026 Product Hunt launch finished #1 Product of the Day with 1,112 upvotes.

Pricing: the 1-credit-equals-1-scrape rule

Every scrape costs one credit with residential proxies, JS rendering and anti-bot bypass included; heavier operations (structured extraction, brand retrieval, enrichment) cost 10 credits, and web search bills 1 credit per result. Failed requests are not billed.

The pricing page's own words are the differentiator: "There are no hidden surcharges for stealth requests. JS rendering, anti-bot bypass, and premium proxies are all included at the standard 1 credit per page." That sentence is aimed squarely at how the incumbents bill. Firecrawl's stealth mode costs 5 credits per request — a 5x multiplier for exactly the hard pages where you need it — and Apify's bill stacks three meters (compute units, proxy gigabytes at $8/GB residential, plus any paid Actor's own fee). Context.dev's bet is that predictability wins.

The credit menu beyond plain scraping: crawls bill 1 credit per page and document parsing 1 per file; screenshots and font extraction cost 5; structured extraction, brand retrieval, styleguides and all five enrichment endpoints cost 10 per call. The one genuinely expensive corner is web search, billed at 1 credit per result — a 10-result search costs 10 credits, where Firecrawl charges 2 credits per 10 results. If search is your dominant workload, that difference compounds fast.

Fine print worth knowing: failed requests aren't billed, annual billing takes two months off, and a startup/nonprofit discount of up to 30% for the first year is listed. One caveat from our checking: Context.dev's docs still describe an older tier lineup (different names, an 800/min top rate limit and $6 overage) that disagrees with the live pricing page — we've used the pricing page throughout, since it matches current checkout.

Context.dev vs Apify vs Firecrawl

Pick Context.dev for one flat bill and bundled brand/entity data, Apify for its 52,000-Actor marketplace and heavy automation, and Firecrawl for browser interaction and the cheapest plain-scraping entry point.

Some arithmetic from the published prices (our math, not any vendor's): Context.dev Pro works out to about $0.75 per 1,000 scrapes with stealth capability included. Firecrawl Standard is roughly $0.83 per 1,000 plain scrapes — but about $4.15 per 1,000 in stealth mode. Apify isn't directly comparable because it bills compute units, which vary with each Actor's memory and runtime; our [Apify pricing explainer](/blog/apify-pricing-explained) walks that model. The honest summary: for hard-to-scrape pages at volume, Context.dev's flat rate is the cheapest predictable number on the table; for easy pages, the three are close.

What the flat rate doesn't buy: Apify's store lists 52,273 prebuilt Actors — ready-made scrapers for specific sites — while Context.dev has none; you write your own calls. Firecrawl's Interact feature drives a real browser (clicking, typing, sessions) at 2 credits per browser-minute; Context.dev has no browser-automation endpoint at all. And this is a crowded field — as one Hacker News commenter put it on the founder's own launch thread, "it's probably one of the most crowded spaces." The founder's answer to the differentiation question was billing plus bundling: "1 credit = 1 scrape, no hidden credit multiplier… we have world class brand data… we're focused primarily on the infra use-case."

A small sign of where this market is heading: Firecrawl's own pricing page now titles itself "the context API to search, scrape, and interact with the web" — the incumbents and the newcomer are converging on the same pitch. For where each tool fits an agent stack today, our [web scraping for AI agents](/blog/web-scraping-for-ai-agents) ranking covers the trade-offs in depth.

The honest caveats before you build on it

A solo-founder company with SOC 2 Type I (vendor-claimed), an SLA only on Enterprise, quarterly brand-data refresh, no marketplace or browser automation, and near-zero independent community track record — promising, not yet proven.

The track-record picture cuts both ways. We found zero Reddit discussion of Context.dev, no G2 or Trustpilot listing, and just eight Product Hunt reviews (4.9/5) as of July 2026 — no complaints anywhere, but no production war stories either. The customer logos on its homepage (Mintlify, daily.dev, and on the YC profile, Klarna) are vendor-stated; we could not independently confirm any of them. The vendor's own numbers also wobble between contexts: "5,000+ businesses" on Product Hunt vs roughly 284 companies on the YC profile — most likely signups versus production customers, but that's our inference, not their clarification.

The Launch HN thread is worth reading for the skeptical case: commenters pushed on market saturation, on differentiation from Firecrawl, and on scraping ethics (one self-hosting commenter's response was simply "great, another thing I have to block server side"). None of that is disqualifying — every scraping vendor faces the same objections — but it is the environment this product competes in.

• Solo-founder YC S26 company (team of one) — fast-moving, but a real bus-factor question for production infrastructure
• SOC 2 Type I only (vendor-claimed; Type II is the one that audits controls over time); 99.9% SLA gated to Enterprise
• Brand data refreshes quarterly, and first-time brand lookups take 10–30 seconds uncached
• Web search bills per result — 10x Firecrawl's search rate at equal result counts
• No prebuilt-scraper marketplace and no browser automation — you write every integration yourself
• Docs and pricing page disagree on tier names and limits; the pricing page is canonical

Who should pick Context.dev — and who shouldn't

Pick it if you want one flat, predictable bill for agent-facing scraping plus brand and entity data through one API or MCP server. Skip it if you rely on prebuilt scrapers, browser automation, or need a long compliance track record today.

The free tier is the right way in either direction: 500 credits a month with a work email, hard-stopping instead of billing overage, is enough to run a real pilot against the pages you actually scrape. If your workload is mostly hard, bot-protected pages, run the same pilot through Firecrawl's stealth mode and compare the two bills — the 5x surcharge gap is the whole argument.

• Pick Context.dev: agent builders who want scraping + search + brand/entity data behind one key and one MCP server, with a bill you can predict from request counts alone
• Pick Context.dev: teams scraping bot-protected pages at volume, where competitors' stealth surcharges dominate the bill
• Skip it: anyone whose workflow leans on Apify's ready-made Actors — see our [web scraping for AI agents](/blog/web-scraping-for-ai-agents) ranking for that trade-off
• Skip it: browser-automation workloads (clicking, sessions, logins) — Firecrawl's Interact or an Apify Actor covers what Context.dev doesn't offer
• Skip it (for now): compliance-heavy production where SOC 2 Type II and a public uptime history are hard requirements

Related ToolProven pages