←News detail

QbitAIUpdated Nov 1 17:29

TypeSafe AI and Treg Open-Source a People-Finding Tool That Is 85% Cheaper Than Clay

AI summary

TypeSafe AI has teamed up with the agent tool platform Treg to open-source a people-finding tool that it says is more than 85% cheaper than Clay and has already gained 6,400 stars. After users connect Treg to an Agent and describe the target person, the Agent breaks down criteria such as region, funding stage, and job title to query a people database, and then Jev determines whether a candidate matches and fills in LinkedIn information and a verified email address. In a comparison test of 292 people, Treg found 289 email addresses, 264 of which matched the companies' publicly listed addresses, at a cost of $1.49; Clay found 280, with 262 matches, at a cost of $10.34. Looking up a single email address costs as little as $0.0089, with no charge if nothing is found, while email verification and résumé completion are billed separately.

Why read this

AI score 65

The original article provides a measured price comparison against Clay from the same batch of listings, along with People Search Bench results, which can be used to assess the cost structure of this type of people-finding tool.

Topics

Tags

Body · AI translation

Good grief, Jev's new results are out, once again the "price-performance butcher"!

This time, they've charged straight into the people-search space.

Just now, Jev's parent company TypeSafe AI teamed up with the Agent tool platform Treg to open-source a people-search tool that is 85% cheaper than Clay, raking in 6,400 Stars.

From now on, if you ever need to find someone on the fly, just tell it to get to work;

for as little as $0.0089 (about 6 cents), it can fish out a useful lead for you.

As soon as the news broke, the comments section erupted.

Some folks, while exclaiming "this is insane," rushed to hook it into their own dashboards and start testing;

others had their pupils quake on the spot at the price:

My god, this price is just absurdly cheap.

And TypeSafe AI, which never misses a chance to meme, whipped out a classic meme from Taken, summing up Jev's core capability perfectly:

I will look for you, I will find you.

I will look for you, I will find you.

All right, all right, it looks like the Agent "I will find you" era is really about to arrive. (doge)

If an Agent wants to find someone, it just has to say the word

The new tool is simple to use: you just connect Treg to an Agent, then tell the Agent who you want to find.

For example:

Please help me find the marketing leads at post-Series B AI startups in the San Francisco area, and fill in their verified work emails.

Once the goal is clearly stated, the Agent starts hunting around for people based on this "persona."

It first breaks out conditions like region, funding stage, industry, and job title, then calls the corresponding people databases through Treg to pull back a batch of candidates who might fit the bill.

But finding names is only the first step.

Next, Jev steps in as the "judgment gate."

Based on the information already found, it judges one by one whether these candidates are really the roles the user is looking for, and kicks out those who "seem related but are actually off track."

After passing the screening, the Agent, via Treg, calls tools for people search, email lookup, and email verification to fill in the candidates' profiles, work history, LinkedIn information, and verified work emails.

Overall, this whole division of labor flows smoothly and with plenty of chemistry;

the Agent handles understanding the task and orchestrating the workflow, Jev handles judging who better fits the requirements, and Treg plugs in the data tools behind the scenes to do the work.

And so, the entire workflow turns, link by link.

In the end, what the Agent hands us is no longer a long list of names requiring manual double-checking, but a list that can be taken straight to recruiting, prospecting, or people research.

Moreover, this people-search workflow no longer has to be trapped inside a new platform.

With a single Treg key, the Agent can search and call among the thousands of tool interfaces it connects to.

So, if you already have accounts with providers like Hunter, you can plug in your own keys directly;

if not, Treg can call on your behalf first, deducting per-use fees from a prepaid balance. (Calls with your own key don't go through Treg billing.)

And just like that, this people-search pipeline gets smoothly taken over by the Agent.

666, in the old days we hand-crafted our way through a pile of tools to find people.

Now, we just say who we want to find, and the Agent can bring them back.

This really has a bit of that Liam Neeson-in-Taken vibe, declaring over the phone, "I will find you."

As for whether it can actually find the right people, Treg also brought out its People Search Bench results.

This benchmark includes 119 real tasks, covering scenarios like recruiting, B2B prospecting, deterministic lookup, and influencer search. Take B2B prospecting: Claude Code going it alone had a task accuracy of 43%, which rose to 78.2% after connecting to Treg.

In the other three categories, Treg scored 80%, 76.3%, and 62.9%, respectively.

Of course, benchmark scores are one thing; whether it can find people still depends on real-world results.

Next, let's dig into what these 6 cents can actually buy.

What exactly can 6 cents buy?

According to Treg, after giving a name, company domain, or LinkedIn profile, for as little as $0.0089 (about 6 cents), you can look up a person's email, and you're not charged if nothing is found.

However, these 6 cents only cover the search.

If you also want to verify whether an email is real, fill in a personal résumé, or check what recently happened at a company, you'll need to pay extra.

Even so, it's still considerably cheaper than traditional subscription tools.

To prove it wasn't bluffing, Treg also went head-to-head with Clay using the same list of names.

The test included 292 people in total.

Treg found emails for 289 of them, with 264 exactly matching the addresses publicly listed by their companies, at a total cost of $1.49 (about 9.93 yuan).

Clay, meanwhile, found 280 email addresses, of which 262 were exact matches, at a total cost of $10.34 (about 68.94 yuan).

△ Image source: Treg

The two sides differed by only two in the number of people correctly found, yet the price gap was enormous.

On average, to correctly find one email address, Treg needed only $0.0056 (about 0.037 yuan), while Clay needed $0.0395 (about 0.263 yuan); by this measure, Treg was over 85% cheaper than Clay.

And now, Treg is also giving every new user a free $1 testing credit.

It looks incredibly tempting!

But even so, cheap as it is, some users have now tried it themselves and found that this new tool still needs more practice when it comes to accuracy:

It's good and all, but when I searched for myself or other teammates, Treg didn't find anything meaningful, and it was even hard to find my LinkedIn.

At the same time, ChatGPT and Gemini did find a lot, though of course they were much more expensive.

In response, Treg didn't try to hide anything; instead, it openly chased after the other side asking for homework:

Can't find it? Send me the test set right away, and we'll tune it on the spot!

And to another group of users worried that "please do it yourself, we don't need a middleman marking up the price," Treg also quickly raised its hand and responded:

Don't worry, we don't take a cut!

Anyway, it looks pretty good; at this price, what more could you ask for? As a ten-level internet surfer, I'm about to start pedaling right away!

Whether it's "I will find you" or "I tried to find you," we also welcome friends to come back and share their experience using it. (doge)

One More Thing

By the way, ever since the decision model Jev exploded overnight in September, TypeSafe AI really hasn't had a moment to rest.

Within just three days, Jev's token usage broke through 1 trillion; TypeSafe AI subsequently claimed that about one-third of Fortune 500 companies had begun using Jev.

Capital also came rushing in after hearing the news.

Just 24 days after Jev was released, TypeSafe AI completed an $870 million Series A funding round, with its valuation soaring 37-fold to $7.5 billion (about 54.375 billion yuan), completely ending its obscure status of the previous two years;

At the same time, Jev had just carved out a new track for "decision models" in front, and right behind it, a group of big companies including Microsoft and OpenAI rushed to the scene carrying shovels.

(For details, see: Jev valued at $10 billion! Founder Diogo Almeida answers everything)

However, TypeSafe AI has never stopped moving forward.

On October 14, Jev 1.14.0 launched, simultaneously adding new capabilities such as image input, longer context, and Box;

△ Image source: TypeSafe AI official website

Now, TypeSafe AI has teamed up with Treg to put judgment capability into the people-finding process.

In this process, matters awaiting handling—products, financing, customers, competitors, community needs—all came crashing in at once.

TypeSafe AI co-founder and CEO Diogo once humorously described himself on a podcast:

Like an empty shell with its soul drained out, busy putting out fires everywhere every day; I really want to get a good sleep!

In just one sentence, it truly captures the bitterness of an entrepreneurial CEO who has to catch the overwhelming traffic from a product going viral, respond to big companies entering the field to besiege them, and also worry about company management. (laugh)

I can only say, I wish Diogo, while continuing to lead Jev into battle and stir things up, can soon realize this simple yet luxurious wish to sleep~

Source

QbitAI