Since common AI agent Manus was launched final week, it has unfold on-line like wildfire. And never simply in China both, the place it was developed by Wuhan-based startup Butterfly Impact. It’s made its approach into the worldwide dialog, with influential voices in tech, together with Twitter co-founder Jack Dorsey and Hugging Face product lead Victor Mustar, praising its efficiency. Some have even dubbed it “the second DeepSeek,” drawing comparisons to the sooner AI mannequin that took the trade without warning—each for its sudden capabilities and its origin.
Manus claims to be the world’s first common AI agent—leveraging a number of AI fashions (resembling Anthropic’s Claude 3.5 Sonnet and fine-tuned variations of Alibaba’s open-source Qwen), and varied independently working brokers to behave autonomously on a variety of duties. (That is totally different from AI chatbots, together with DeepSeek, which can be based mostly on a single giant language mannequin household and are primarily designed for conversational interactions.)
Regardless of all of the hype, only a few folks have had an opportunity to make use of it. Presently, below 1% of the customers on the waitlist have obtained an invitation code. (It’s unclear how many individuals are on this waitlist, however for a way of how a lot curiosity there’s, Manus’s Discord channel has greater than 186,000 members.)
MIT Know-how Evaluation was in a position to receive entry to Manus, and once I gave it a take a look at drive, I discovered that utilizing it seems like collaborating with a extremely smart and environment friendly intern: Whereas it often lacks understanding of what it’s being requested to do, makes incorrect assumptions, or cuts corners to expedite duties, it explains its reasoning clearly, is remarkably adaptable, and might enhance considerably when supplied with detailed directions or suggestions. Finally, it’s promising however not good.
Identical to its father or mother firm’s earlier product, an AI assistant known as Monica that was launched in 2023, Manus is meant for a world viewers. English is about because the default language, and its design is clear and minimalist.
To get in, a consumer has to enter a legitimate invite code. Then the system directs customers to a touchdown web page that intently resembles these of ChatGPT or DeepSeek, with historic periods displayed in a left-hand column and a chat enter field within the heart. The touchdown web page additionally options pattern duties curated by the corporate—starting from enterprise technique improvement to interactive studying to personalized audio meditation periods.
Like different reasoning-based agentic AI instruments, resembling ChatGPT DeepResearch, Manus is able to breaking duties down into steps and autonomously navigating the net to get the data it wants to finish duties. What units it aside is the “Manus’s Pc” window, which permits customers not solely to look at what the agent is doing, but additionally intervene at any level.
To place it to the take a look at, I tasked Manus with three assignments: (1) compile an inventory of notable reporters masking China tech, (2) seek for two-bedroom property listings in New York Metropolis, and (3) nominate potential candidates for Innovators Under 35, an inventory created by MIT Know-how Evaluation yearly.
Right here’s the way it did:
Job 1: The primary record of reporters that Manus gave me contained solely 5 names, with 5 “honorable mentions” beneath them. I observed that it listed some journalists’ notable work whereas not others. I requested Manus why it did this. The explanation it supplied was hilariously easy: It obtained lazy. It was “partly as a consequence of time constraints as I attempted to expedite the analysis course of,” the agent advised me. After I insisted on consistency and thoroughness, Manus responded with a complete record of 30 journalists, noting their present outlet and itemizing notable work. (I used to be glad to see I made the minimize, together with a lot of my beloved friends.)
I used to be impressed that I used to be in a position to make top-level recommendations for adjustments, a lot as somebody would with a real-life intern or assistant, and that it responded in form. And whereas it initially missed some journalists’ employer standing adjustments, once I requested it to revisit some outcomes, it rapidly corrected them. One other good function: the output was downloadable as a Phrase or Excel file, making it simple to edit or share with others.
Manus hit a snag, although, when accessing journalists’ information articles behind paywalls; it regularly encountered CAPTCHA blocks. Since I used to be in a position to comply with alongside step-by-step, I may simply take over to finish these, although many media websites nonetheless blocked the device as a consequence of suspicious exercise. I see potential for main enhancements right here—and it will be helpful if a future model of Manus may proactively ask for assist when it encounters these kinds of restrictions.
Job 2: For the condo search, I gave Manus a fancy set of standards, together with a finances, and parameters for a spacious kitchen, outside area, entry to downtown Manhattan, and a significant prepare station inside a seven-minute stroll. Manus initially interpreted obscure necessities like “some form of outside entry” too actually, utterly excluding properties and not using a personal terrace or balcony entry. Nevertheless, after extra steerage and clarification, it was in a position to compile a broader and extra useful record, giving suggestions in tiers and neat bullet factors.
The ultimate output felt straight from Wirecutter, containing subtitles like “greatest general,” “greatest worth,” and “luxurious choice.” This activity (together with the backwards and forwards) took lower than half an hour—loads sooner than compiling the record of journalists (which took slightly over an hour), seemingly as a result of property listings are extra brazenly obtainable and well-structured on-line.
Job 3: This was the most important in scope: I requested Manus to appoint 50 folks for this 12 months’s Innovators Beneath 35 record. Producing this record is a gigantic endeavor and we usually get lots of of nominations yearly. So I used to be curious to see how nicely Manus may do. It broke the duty into steps, together with reviewing previous lists to know choice standards, making a search technique for figuring out candidates, compiling names, and guaranteeing a various choice of candidates from everywhere in the world.
Growing a search technique was probably the most time-consuming half for Manus. Whereas it didn’t explicitly define its strategy, the Manus’s Pc window revealed the agent quickly scrolling by web sites of prestigious analysis universities, bulletins of tech awards, and information articles. Nevertheless, it once more encountered obstacles when attempting to entry tutorial papers and paywalled media content material.
After three hours of scouring the web—throughout which Manus (understandably) requested me a number of instances whether or not I may slender the search—it was solely in a position to give me three candidates with full background profiles. After I pressed it once more to supply a whole record of fifty names, it will definitely generated one, however sure tutorial establishments and fields had been closely overrepresented, reflecting an incomplete analysis course of. After I identified the difficulty and requested it to search out 5 candidates from China, it managed to compile a strong five-name record, although the outcomes skewed towards Chinese language media darlings. Finally, I had to surrender after the system warned that Manus’s efficiency may decline if I stored inputting an excessive amount of textual content.
My evaluation: Total, I discovered Manus to be a extremely intuitive device appropriate for customers with or with out coding backgrounds. On two of the three duties, it supplied higher outcomes than once I introduced the identical duties to ChatGPT DeepResearch, although it took considerably longer to finish them. Manus appears greatest suited to analytical duties that require in depth analysis on the open web, however have a restricted scope. In different phrases, it’s greatest to stay to the kinds of issues a talented human intern may do throughout a day of labor.
Nonetheless, it’s not all clean crusing. Manus can undergo from frequent crashes and system instability, and might battle when requested to course of giant chunks of textual content. The message “As a result of present excessive service load, duties can’t be created. Please attempt once more in a couple of minutes” flashed on my display screen a couple of instances when beginning new requests, and sometimes Manus’s Pc froze on a sure web page for a protracted time frame.
It has a better failure charge than ChatGPT DeepResearch—an issue the group is addressing, according to Manus’s chief scientist Peak Ji. That stated, Chinese language media outlet 36Kr experiences that Manus’s per-task price is about $2, which is simply one-tenth of DeepResearch’s price. If the Manus group strengthens its server infrastructure, I can see the device changing into a most well-liked selection for particular person customers, notably white-collar professionals, unbiased builders, and small groups.
Lastly, I feel it’s actually priceless that Manus’s working course of feels extra clear and collaborative. It actively asks questions alongside the best way and retains key directions as “data” in its reminiscence for future use, permitting for an simply customizable agentic expertise. It’s additionally very nice that every session is replayable and shareable.
I count on I’ll maintain utilizing Manus for all kinds of duties, each in my private {and professional} lives. Whereas I’m unsure the comparisons to DeepSeek are fairly proper, it serves as additional proof that Chinese language AI corporations should not simply following within the footsteps of their Western counterparts. They don’t seem to be simply innovating on base fashions, however actively shaping the adoption of autonomous AI brokers in their very own approach.
