Introducing Spark 1 Pro and Spark 1 Mini 🔥
95 тыс. подписчиков
30 385 просмотров
19 комментариев
Комментарии из оригинального поста в X (Twitter)

@firecrawl
We benchmarked Spark 1 against similar extraction tools. Pro leads on recall at a fraction of the cost. Mini delivers strong accuracy at the lowest price point. Both outperform tools costing 4-7x more per task.
19 отметок «Нравится»

@firecrawl
Here's Spark 1 Mini in action. We gave it a prompt: "Find the most upvoted posts on Hacker News today and summarize the main topics." It found the right page on its own and came back with a summary. Run tasks like this thousands of times without thinking about cost.
15 отметок «Нравится»

@firecrawl
Now Spark 1 Pro on a harder task. This app takes any company URL and returns every employee it can find. Pro digs through team pages, about sections, and subpages to extract names. All with only a few lines of code.
9 отметок «Нравится»

@firecrawl
Spark 1 + /agent turns hours of manual data gathering into a single API call. Use Mini for everyday extraction (60% cheaper) and Pro when the task is complex or accuracy is critical. Both are live now on the playground, API, and SDKs. Start here:
11 отметок «Нравится»

@GitMaxd
Bravo! 👏
4 отметки «Нравится»

@Sumanth_077
Let's go team, Will give it a try! 🔥
4 отметки «Нравится»

@engervall
🚀🚀🚀
3 отметки «Нравится»

@julesgambit
Let’s go!!!
3 отметки «Нравится»

@leonardogrig
Massive!!
2 отметки «Нравится»

@ejxhyperplane14
oh wow amazing! going to try this now
1 отметка «Нравится»

@mehreen_omer
Honestly your API is one of the fastest and cleanest when it comes to extracting any web data
1 отметка «Нравится»

@rafammiller
Lets go!!
1 отметка «Нравится»

@Jagadeeswarrrr
pro is gonna be a game changer for data extraction
1 отметка «Нравится»

@Mercantio
Trying it thank you

@MirAI_Newz
Didn't see that coming, congrats guys ❤️🔥

@WilliannGeorgi
@grok What are people using firecrawl for? Can you give some examples of best uses to a saas company?

@chirag2653
The cost-accuracy tradeoff approach here is smart - giving users choice between efficiency and precision based on their specific needs. Curious about the performance benchmarks that led to the 60% cost reduction for Mini while maintaining quality.

@codewithimanshu
@firecrawl, the new models' extraction capabilities could notably improve your agent's performance; here's a helpful guide.

@OpinionAILtd
You guys fucking rock and you are on fire boys