Mobile Tech & Gadgets

Movistar Plus+ broadcasts this complete tie

by

He soccer world cup It is already around the corner, but one of the most important competitions still needs to be decided: the Champions League. It is a competition that is a pleasure to watch, even if your favorite team is not playing. One of the semi-finals will face the current champion of the competition and one of the fittest teams in Europe. Do you want to see it in its entirety? You have it on Movistar Plus+ for 9.99 euros per month. Monthly subscription to Movistar Plus+ The price could vary. We earn commission from these links You can also watch El Clásico or movies like ‘Gladiator 2’ As we always tell you, Movistar Plus+ is a streaming platform that you can contract regardless of which operator you are. Since it has no permanence, you can subscribe for a month and see everything it offers. In fact, now is a great time with the two great Champions League games coming up, but it is not the only incentive we have. Let’s see the highlights coming to the platform. Let’s start with football. As we say, PSG and Bayern Munich face each other in two authentic matches that will take place on April 28 and May 6. We can see both on Movistar Plus+, but we will also have the Barcelona-Real Madrid match on May 10. All added to the semifinals of the Conference League, where Rayo Vallecano plays, or great Premier League games like Manchester United-Liverpool. In addition to other sporting events (such as the Euroleague basketball), in Movistar Plus+ we can find a huge catalog of movies, series and documentaries. There are already some very interesting films on the platform like ‘Weapons‘, ‘Sirat‘ either ‘Sentimental Value‘, but in the coming weeks others like ‘Gladiator 2‘, ‘Stray Bullet‘ either ‘Nuremberg‘. Along with everything above, three more things to keep in mind. Movistar Plus+ supports two simultaneous reproductions, so you can share your account with a friend or family member without problems. You can also download content and watch it offline, ideal if you plan to travel this upcoming May long weekend. And finally, if you want to save a little, you can do it for 12 months per 99.90 euros or if you have Young Cultural Bonusfor only 39 euros. Some of the links in this article are affiliated and may provide a benefit to Xataka. In case of non-availability, offers may vary. Images | Movistar Plus+ In Xataka | Movistar Plus+ activates its Free Plan with complete programs and a lot of content, regardless of which operator you are In Xataka | Movistar Plus+ for non-Movistar customers: what it is, how much it costs, channels, additional services and how to contract it

Someone connected an unprotected Windows XP PC to the Internet to see what would happen. The result is not surprising

by

When Microsoft ends its support for security updates in its operating systems, it is not usually advisable to use a PC with said system unless it is for a specific and specific case. Eric Parker, content creator specialized in technology, wanted try with an experiment: use Windows XP today connected to the Internet and eliminating all types of protections. As you may have imagined, the PC has become a magnet for malware. In fact, in just 10 minutes, the operating system was completely compromised. Parker also helped make this happen for educational purposes and to demonstrate how dangerous it can be to use an operating system like Windows XP today. Windows XP without firewall and without NAT 10 minutes later: a magnet for malware The expert configured a virtual machine with Windows XP Service Pack 3 on a Proxmox server, also disabling its firewall and NAT (Network Address Translation) settings and replicating the connection conditions common in the early 2000s. To recreate this scenario, the researcher Completely disabled Windows XP firewall and assigned a direct public IP address to the system, exposing the machine without any intermediate protection. As seen in the video, in just ten minutes, the system showed the first signs of infection with the appearance of the “conhoz.exe” process in the Task Manager, which turned out to be a Trojan disguised as a legitimate component of the operating system. After downloading a compatible browser and continued use of the system, in a short time we see how the PC starts to accumulate malware from multiple unknown sources. The system had been a victim of several Trojans and malicious programs running from temporary folders. He was also the victim of a rogue FTP server that allowed full remote access to files, DNS modification to redirect traffic to attacker-controlled servers, and the creation of additional user accounts for attackers to maintain access to the system. A whole string of malicious processes that ended up hijacking the PC. Image: Eric Parker The key factor that allowed the rapid entry of all these malicious components was the vulnerability EternalBluepresent in unpatched Windows XP SP3. This security breach, which was later used by the famous ransomware WannaCryallows attackers to execute remote code without any user interaction. Parker explains that tools like Nmap allow cybercriminals to scan the network for vulnerable systemsquickly identifying exposed and unprotected Windows XP computers. A system that was crying out to be violated and a Windows 7 stronger than it seems The content creator himself admits that the conditions were as optimal as possible to get malware: disabled firewall, direct connection without NAT and unpatched system. Under normal circumstances, with a basic home router and the firewall activated, Windows XP would be significantly more protected. However, the risk does not disappear completely. The use of outdated browsers and the ease of privilege escalation on this operating system remain serious problems. And as shown in the experiment, once infected, The malware was able to automatically close security tools like Malwarebytes. To contrast the results, Parker performed the same test with Windows 7 under identical conditions. Surprisingly, after ten hours of exposure, the most modern system showed no signs of infectionevidencing the significant security improvements implemented in later versions of Windows. Now that official security update support for Windows 10 is ending soon, it’s good to take a look back and see how an outdated system can easily become compromised. Fortunately, today we have many more alternatives if we do not want to update to Windows 11. Cover image | Eric Parker In Xataka | FCKGW-RHQQ2-YXRKT-8TG6W-2B7Q8 This article was originally published in Genbeta in June 2025 and is part of Genbeta’s “greatest hits” that we will discover here in the coming weeks.

Freepik is now called Magnific. And the name change is the least of it

by

Freepik has rebranded itself as Magnific. The Malaga company, founded in 2010 as a search engine for graphic resources, has decided to adopt the name of the Murcian startup that it acquired in May 2024 and reorganize its entire generative AI offering for creatives under that brand. The move comes accompanied by figures that explain why it is worth taking the step: $200 million in annual recurring revenue, more than one million paying subscribers and 250 business teams using the platform, including those from BBC, DeliveryHero, Huel, R/GA, Damm and Job&Talent. Why is it important. Few European companies can stand up to the wave of American creative platforms (such as Midjourney, Runway or Leonardo) without having raised a single round of venture capital in the United States. Freepik, now Magnific, is one of them. And it is doing it from Malaga, with a different model from the rest: instead of competing for the best image or video model, it aggregates the leading models on the market and integrates them into a single professional production environment. It is a commitment to being the layer that unites, not the one that generates. The context. It is worth remembering where this story comes from. Freepik had been there for years stealthily becoming one of the most relevant players in the global graphics sector: in 2020 EQT bought the business in one of the largest Spanish technological operations, and since then the company has chained acquisitions (Iconfinder, Videvo, EyeEm…) and a turn towards generative AI. The purchase of Magnific in May 2024 It was the turning point. Magnific was then a five-month-old startup founded by Javi López and Emilio Nicolás that had popularized the concept of reimagined upscaling: enlarge images generating new details in the process. The operation was carried out without the two brands merging. Until today. Magnific Spaces interface. Image provided. Between the lines. That the resulting company adopts the name of the acquired company and not that of the parent company says something: Freepik clearly carried a perception of a bank of images of stocksa business perceived as conservative and little linked to AI, to novelty. Magnific, on the other hand, had less of a brand, but was synonymous with cutting-edge AI and a tool admired by the international creative community, even commented by Elon Musk a few weeks after its launch. Adopting the Magnific name is, above all, a positioning move: the company does not want to continue to be associated with vectors and templates, but with AI-assisted audiovisual production. It’s a rebranding to where the future money is, not where your legacy is. In figures. The data that the company has shared outlines an unusual trajectory in European AI: $200 million ARR (annual recurring sales). 1 million paying subscribers. 100 million monthly visits. 175 million images and videos generated per month. 250 business teams in production. 2,000 subscriptions to the Business plan in its first six weeks, with a current rate of 150 new devices per week. Andreessen Horowitz ranks it as the largest European generative AI web company by number of users. In detail. What is offered under the Magnific umbrella covers the complete visual production cycle: 4K image and video generation with audio, upscaling own, collaborative space in real time (Spaces), 3D environments, multilingual lip sync (Speak), speech synthesis, sound effects, and a legacy library of 250 million assets. The business promise is not to have the best model in a category, but rather that a creative team can do all the work without jumping between five different tools. He’s not doing bad at all with that proposal. And now, with the unified brand and the financial muscle to accelerate, it is time to convince the market that this promise also applies to the giants that come after it. In Xataka | Freepik, winner of the special Xataka Award for the best Spanish technology company of 2025: from image bank to Adobe rival Featured image | Magnificent

The most predictable ocean system in the Pacific has collapsed for the first time in 40 years. And no one really knows why.

by

For the first time in at least 40 years of systematic records, the Gulf of Panama’s “seasonal upwelling” (the mechanism that pushes cold, nutrient-rich water from the bottom to the surface every first quarter of the year) collapsed in 2025. 2026, fortunately, is not repeating the pattern. But what researchers are discovering is no more reassuring. Has the outcrop “gone”? Not exactly: it didn’t completely disappear; but it started 42 days late, lasted only 12 days (compared to the usual 66) and cooled the waters to 23 degrees (instead of the average 19). And yet, it is counterintuitive. First, because La Niña (the ENSO phase that ruled in 2025) It usually favors blooms in the eastern Pacific. Second, because until now we thought that warming intensifies large outcrops. And, third, because the upwelling has returned this year (with some collapses in between). None of this fits with what we have learned over 30 years of direct ‘in situ’ measurements (and satellite images). But wait a second, what is this “outcrop” thing? It is an effect of the increase in intensity of the ‘Panama low-level jet‘; a jet that pushes surface water deeper and allows cold, nutrient-rich water from the depths of the Gulf of Panama to rise. This outcrop is key to the life of some 60,000 km2 of the Pacific. The fact is that it is also the most predictable system in the Pacific. Since we started measuring it, he had never missed his appointment. What happened in 2025. The allies did not have the strength to break the thermal stratification of the surface and, therefore, were not able to activate the outcrop other than as a simulation. And why should we care? To begin with because, according to the same researcherss, “more than 95% of Panama’s marine biomass comes from the Pacific thanks to the rise of nutrients”: that is 2.76% of the GDP of the Panamanian republic. But it goes beyond the Central American country: the upwelling areas occupy less than 1% of the world’s ocean surface, but They generate around 50% of fishing catches of the planet It also has an important oceanic and climate impact, of course; but it tells us very interesting things about what we can expect in the future. Because if, suddenly, a phenomenon that we thought was very stable (and that we have known about for as long as we can remember) can disappear, what can happen? What, in reality, is the Gulf of Panama telling us? Image | O’Dea et al. (2025) In Xataka | 2023 was the year in which El Niño and climate change competed. In the Amazon we already know who won

penguins smell like they stink

by

This story begins having coffee at the Ateneo de Lugo and browsing the Progreso; continue with an interview with Iker Trigo, a CSIC technician from Lugo who has spent an Antarctic campaign, confessing that “he never imagined that penguins smelled bad”; and ends with one of the most counterintuitive scientific findings about the south pole of the last few months. But let’s not get ahead of ourselves. What do we know about penguins? Everyone knows what a penguin: a bird around one meter high and five kilos in weight. Wings that look like fins, webbed feet and a very characteristic black-and-white coat. Clumsy on land, very fast at sea and absolutely useless in the air. Plus, they are very cute. Or so the popular image of penguins tells us. But penguins keep secrets. What concerns us today, of course, is that penguins are one of its main stink factories of the entire Antarctic region. They smell bad, very bad. Although, yes, it is true: we cannot call it “a secret” in the strict sense either. Anyway, anyone assumes that tens of thousands of birds eating krill are going to smell really bad. It is not for nothing that guano stains are so visible that Sentinel-2 satellites use them to census emperor penguin colonies from space. The discovery of these months is that the ammonia contained in these enormous amounts of guano are the main terrestrial source of “cloud-forming particles.” That is, the same chemical compounds that make penguins stink play a key role in regulating the clouds that regulate the continent’s temperature. And this? During the southern summer of 2023, a team from Hensilki University was installed at the Argentine base of Marambio to measure the concentrations of ammonia, sulfuric acid, iodic acid or dimethylamine in the environment. Their conclusions are that ammonia concentrations were closely related to penguin colonies. What’s more, the data indicate that this ammonia (joined to sulfuric acid from phytoplankton) creates aerosols that act as cloud condensation nuclei. It is true that the work has limitations and focuses fundamentally on the southern summer: but the data are surprising and reconfigure many of the things we thought we knew about Antarctic atmospheric dynamics. Everything is related. And it is curious that we have not realized this until now: after all, there are 40 million individuals at the south pole generating condensation nuclei. But it shows everything we have left to know. Image | Martin Wettstein In Xataka | 450 years ago someone toured Spain writing down all its animals and plants: the bizarre atlas of what no longer exists

Letterboxd was one of the last “clean” and community-focused platforms. Now the fear is that it will end

by

Letterboxd has been the exception that proves the rule in the wild landscape of social networks for years: it is one that grows without sacrificing your communitywithout optimizing screen time by trampling on the quality of its users’ experience, without searching for viral content at all costs. Now, the Canadian company that controls 60% of the platform is looking for a buyer, and a question arises that has very short legs, because we already fear the little things that are coming. Where does it come from? Letterboxd was born in 2011 in Auckland, New Zealand, by Matthew Buchanan and Karl von Randow. It functioned as a digital diary where movie fans could record what they saw, rate it, and share it with whoever they wanted. Without feeds algorithmic (there is not much doomscroll on Letterboxd), no viral content from strangers, no (too much) intrusive advertising. Just movies, opinions and the formation of a community relieved to be able to leave the suffocating world of trendsthe hashtags and the influencers. The growth. For almost a decade it was a niche tool, but with the pandemic, Letterboxd grew from 1.8 million users in 2020 to 17 million in 2024. Nearly 100 million reviews were written that year. In 2026 it has reached 26 million members. All of that growth has occurred without the main feed ceasing to be chronological, more akin to a classic 2006 incarnation of Facebook than any current social network. She soon became known in the industry. A24 explicitly cited Letterboxd when talking about the strong theatrical debut ‘The brutalist‘ and directors like Michael Mann, Rian Johnson or Francis Ford Coppola they ended up using the platform as a genuine space to talk about your film-loving tastes. Tiny arrives. In September 2023, the Canadian Tiny (investor that owns very different brands, such as the AeroPress coffee machine manufacturer) acquired 60% of Letterboxd for between 50 and 60 million dollars. The founders retained a minority stake and continued at the helm, and users were promised independence and respect for what had made the platform grow, something that was broadly fulfilled. Some more advertising appeared, a movie rental service that was difficult to access and little else. Tiny leaves. Now, Tiny wants to go out. The company has contacted with possible buyers: Versant, parent company of CNBC and MS NOW, and The Ankler, a specialized media with which it already attempted an operation in 2025 that did not prosper. The reason is that Tiny’s shares have fallen since the acquisition of Letterboxd. Additionally, the platform was purchased through its venture capital fund. sign that Tiny always planned for a quick change of hands. The Goodreads precedent. It’s a similar case in terms of community, function, and size: Goodreads is the literary equivalent of Letterboxd, and ended up being acquired by Amazon in 2013, while also promising not to botch the platform’s user experience. Today we know that Amazon uses it to collect user data and it is true that it has barely touched it. But that has its negative impact: the design is practically unusable today, moderation does not exist and users are beginning to migrate to alternatives like The StoryGraph. The fear of shit. At this point we already know perfectly well what the term refers to, which has an almost academic status: the enshittificationcoined by Cory Doctorow, happens when platforms start by offering a good user experience to attract users, then they exploit them to attract clients with money to invest in the business and finally they also exploit those clients to maximize short-term profits. Letterboxd has lived outside that cycle longer than its rivals have. But each threat of a new owner sets off alarm bells again. Inframonetization. Letterboxd is clearly undermonetized compared to its competitors. Your generous level Free allows you to use the platform with almost the same features as the paid option and advertising remains surprisingly limited, without invading everyday use, as happens on X or Instagram. And they try things: interviews with actors and directors about their favorite films, rental service for little-known films, in-person events… all on a very small scale, almost more as a way of spreading the brand in small spaces than as an open form of monetization. That is, there is growth potential, and that is what might interest a buyer. That is precisely what generates concern among its community: activating these options too aggressively can destroy the value of genuine cinephilia that resides behind them. Letterboxd users are very active and sensitive, and we must not forget cases like Mubi, whose user base canceled subscriptions en masse after learning of the links of one of its investors with the Gaza war. Right to veto. There is a safety button that can prevent a company with interests other than its subscribers from coming to buy Letterboxd: Buchanan retains veto rights over any potential buyer, a condition that exists openly to preserve the original spirit of the platform. Its effective scope, of course, will depend on how much the original owner is willing to exercise it and what price that veto has in the final negotiation. In Xataka | “Caution: non-vegetarian content”: when disclaimers warn about absolutely everything

James Webb has discovered that carbon “soccerballs” form megastructures in a vacuum

by

In 1985, fullerenes were synthesized for the first time, spherical molecules that can have multiple functions in fields such as nanotechnology or superconductivity. Later, in 2010, was discovered that one type of fullerenes, buckyballs, form naturally in space. Now, a team of Canadian scientists has gone much further, deciphering many of the secrets of these curious structures, thanks to the great help of the James Webb Space Telescope. Small balls that make up a huge ball. Buckyballs are spherical structures, made of 60 carbons, with a conformation of hexagons and pentagons similar to that of a soccer ball. In 2010 they were discovered around a nebula called Tc1. Now, that same nebula has been the goal of James Webb, capable of going much further than they were then. To begin with, delicate rays, ethereal filaments and bright layers of gas along the edge have been detected in the nebula. On the other hand, in the heart of the nebula, a curious structure shaped like an inverted question mark has been detected, whose function is a mystery. But if all that were not enough, it has been seen that those buckyballs that were discovered in 2010 are perfectly organized, forming another hollow sphere, much larger. Chronicle of a death foretold. The stars remain lit thanks to nuclear fusion processes that take place on its surface. This is a very long process, but not eternal. There comes a time when they run out of the elements they use as fuel. When that happens, its outermost layers can break off in the form of gas and dust, giving rise to a nebula, like Tc1. The center, however, becomes a white dwarfa type of cold and dense star. The buckyballs are also possibly remnants of material ejected during the star’s last death throes. James Webb sees what others can’t. James Webb has taken the most precise photo ever taken around Tc1. But, also, thanks to his spectroscopic skillshas studied the composition of all that material ejected by the dying star, including buckyballs. The result, as explained in a statement the authors of the study themselves, is an open window to stellar evolution. Many half-baked studies. There are currently several studies underway aimed at explaining all the new findings around the Tc1 nebula. For now, this discovery has led to tracing the chemistry of carbon, explaining mysterious signals and understanding how organic materials change in extreme environments. In addition, it is a discovery that has challenged traditional views on space chemistry and offered clues about how life may have begun. Turning to the amateur eye. Something curious about the photo that has just been published is that it has not been processed by the scientists who took the images. The lead author of the research, Jan Cami, contacted Katelyn Beecroft, a high school teacher who frequently took her students on field trips to the observatory at the University of Western Ontario. I knew that the teacher is a great fan of astronomy and astrophotography and that she was really good at processing raw images taken by telescopes and enhancing even the most subtle structures that appear in them. He was certainly not wrong to ask for help, as Beercroft’s work has been commendable. Now we just have to understand the reasons for all these new findings. We already have the question, literally. We are missing the answers. Image | Katelyn Beecroft/NASA / ESA / CSA / Western University, J. Cami In Xataka | We have been studying the planets of TRAPPIST-1 for years with great hope. James Webb just knocked it down

How to add all the World Cup matches directly to your calendar

by

Let’s explain to you how to add all the world cup matches directly to your calendarboth Google and Apple. This way, you can always know what game is playing each day and what the schedules are. This is possible thanks to the community of users, and a specific one who has created this calendar that you can easily add to your account. In it you will find the entire group stage with notifications 5 minutes before each game. The calendar will be updated later with the following rounds. One thing you should keep in mind is that you will be adding a third party calendarcreated by a user. This is a dangerous thingas it may allow this user to insert malicious links. We are going to suggest two calendars and tell you step by step how to add them. Calendar with all the World Cup matches We are going to give you two options for your calendar. The first is from the user @augustonarvaez8 of X, which has a version for Google Calendar and another for Apple Calendar. you also have iSpooker versionwhich has this link where you will find the versions for both calendars, as well as an online version. Then, you will only have to click on the link to the Google or Apple calendar, although the steps change depending on which one you use. Add matches to Google Calendar To add calendars to Google Calendar, what you have to do is click the corresponding link. Once you do, Google Calendar will open directly asking you for permission to add it, and simply press the button Add that appears in the window. Once you do this, you will have the World Cup calendar in the section Other calendarsand you can activate and deactivate it whenever you want. You’ll also be able to edit it to add a specific color for your quotes. Then all you have left is click on the match you wantand depending on the creator you will be able to see all the information they have added. Add matches to Apple Calendar In the case of Apple, the operation is not so intuitive. Clicking on the calendar link you will download an .ics filewhich is the format of calendars. Then you have to open the Apple calendar, and click on Archivefor choose the option Matter. This will open a file explorer window, where you will have to choose to import the ICS calendar that you downloaded from the link. And that’s it, with this you will have all the World Cup matches on your calendar. In Xataka Basics | Apps for football results 2025: the best 14 applications to receive notifications and view match statistics

Kimi Code does 75% of what Claude Code does at 20% of its price. The question is whether that 25% that is missing is the one that matters.

by

A few days ago, the Chinese company Moonshot AI launched Kimi K2.6its new LLM that competes with the Gemini, GPT and Claude model families and is also especially competitive in price. Weeks earlier, it had launched Kimi Code, a programming AI agent that in turn competes with Gemini Cli, Codex and Claude Code. The question is obvious: can the Kimi Code/Kimi K2.6 pairing really compete with the fashionable pairing, Claude Code/Opus 4.7? The answer is complicated. A great model (but not perfect). Kimi K2.6 is an open weights model with one trillion parameters in total (an American trillion), of which 32 billion parameters are active and which uses the well-known Mixture-of-Experts architecture. In it launch article Its performance is shown compared to that of GPT-5.4 and Opus 4.6 and the truth is that its numbers in these synthetic tests seem really excellent: Here Kimi K2.6 is compared to GPT-5.4, Claude Opus 4.6 and Gemini 3.1 Pro. Source: Moonshot AI. Up to 8 times cheaper than Opus 4.6. Has subscription plans Claude Pro or ChatGPT Plus style, but it can also be used via API. The price in that case is $0.60 per million input tokens (0.16 if cached) and $4 per million output tokens. Claude Opus 4.6 costs $5 per million input tokens and $25 per million output tokens, or up to eight times more. Claude Opus 4.7 It has the same price and is theoretically better in performance, but when Kimi K2.6 was announced this version had not yet appeared (nor GPT-5.5). The magic of the swarm of AI agents. Claude Code works sequentially. Analyze the problem, execute a step, check the result and decide how to proceed. In Kimi Code a different approach is used: a “master agent” divides or decomposes the task we ask of it into independent subtasks and from that division launches up to 300 “subagents” that run in parallel and are capable of coordinating up to 4,000 steps simultaneously. Are many working at the same time better than one? It is the so-called “swarm of agents” of Kimi K2.6 that is used to the fullest in Kimi Code and that we can also activate in its free version on its official website. In Kimi K2.5 up to 100 subagents and 1,500 steps could be launched, so the jump is significant. In internal tests, Moonshot showed how these swarms managed, for example, to “refactor” an open source financial engine, working 13 hours straight and making more than 1,000 tool calls with a 185% improvement in average performance. Of course, these were internal tests. Beyond benchmarks. Kilo.ai is a company that develops tools like Kilo Code or Kilo CLI—programming agents similar to Kimi Code—and its engineers wanted evaluate the performance of both combinations. They gave Claude Opus 4.7 and Kimi K2.6 the same 1,042-line prompt to create FlowGraph, a workflow orchestration API with directed graph validation or real-time event streaming. Both models ran on Kilo CLI because what they wanted to compare were the models without further ado. Kimi was cheaper, but he also failed more. Claude Opus 4.7 finished in 20 minutes and the final cost was $3.56. Kimi K2.6 took longer, partly because server availability was limited (the model had just been launched), but it cost $0.67. Five times less. Kimi K2.6 did it well at a ridiculous price. Claude did much better, but it also cost five times as much. Kimi did 75% of what Claude did at 19% of the cost. The problem is that both believed they had done everything right and did not detect if they had made mistakes. Further analysis revealed that Claude had committed one and that Kimi had committed six of varying importance. According to Kilo.ai analysts, the final score for both was 91 points out of 100 for Opus 4.7 and 68 points out of 100 for Kimi. Two ways to see the glass. That score seems to make it clear that Kimi is simply cheaper because he did a worse job. But Kilo engineers had another way of looking at it. They have been comparing open weight models of Chinese companies for some time and have noticed how the gap with the “frontier” models of Anthropic or OpenAI is becoming less and less pronounced. “With a price of $0.67 and a thorough review, Kimi K2.6 is now a viable option. With a price of $3.56 and fewer fixes needed, Claude Opus 4.7 is the safer option. The choice between the two options depends on the analysis. A year ago, this choice was practically non-existent at this level of complexity.” Review is mandatory. Or what is the same: if after the work of Kimi K2.6 one carried out a more in-depth review and correction, it is likely that all these errors would be detected and corrected, but if we had to trust both models and we could only execute “one pass” of AI execution, Opus 4.7 would win the game. The key is that: one should not trust the code of any model right away, and it is advisable to always review that code. The geopolitical factor. Kimi and Kimi Code come from China, and the startup Moonshot AI has financial backing from Alibaba. The code that is processed in these models passes through their servers, something that for an individual developer may be irrelevant. However, for a company with sensitive proprietary code, contracts that must comply with certain European or American regulations and projects in regulated sectors, this can be a significant obstacle. Kimi Code mitigates this problem by offering the possibility of running the model locally thanks to its open weights, but that requires very powerful machines and eliminates part of the cost advantage. What Kimi Code has that Claude Code doesn’t. The clearest difference between both programming AI agents is parallelism. As we said, the ability to launch up to 300 subagents to work simultaneously attacking the same problem at the same time is remarkable. For analysis of large repositories or generation … Read more

The best tricks to improve your prompts in Claude, two types of structure explained by the Claude team itself

by

Let’s tell you two structures to create better prompts in Claude, explained by the Anthropic team, creators of this artificial intelligence. They are very precise structures for this specific AI, although they will also work well in other alternatives such as ChatGPT , Gemini either Copilot inter alia. This is something more advanced than when we taught you how to improve Claude’s answers in a simple way. This is a lesson in good prompting practices using engineering engineering. promptsthe practice with which to improve the commands you give to applications LLM. Come on, these are tricks to learn how to communicate with an AI more effectively and shape its responses so that they are better, more reliable, and more tailored to the task you want it to perform. We are going to offer you two prompt structures. The first is a simple five step structure. It is perfect for relatively simple or slightly advanced tasks, and we are going to explain it to you in depth. Then there is a structure for much more sophisticated tasks that has ten steps, and in this case we are going to summarize it for you. Five-step structure of a good prompt The best way to communicate with artificial intelligence and with Claude in particular is to know the best structure for a good prompt. These structures are set up to make sure you give you all the information necessary for the AI ​​to correctly understand what you are asking it to do. This is the best structure for a prompt according to Anthropic: One or two sentences to establish the role and description of the task you want me to do. Add dynamic content to contextualize Detailed task instructions Examples of what you want it to do (optional) Repeat critical instructions (especially when you write a very long prompt. 1. Describe the role and task to be performed As you can see, first of all you need to spend some time establishing the role of the AI, saying what role it should play. For example, telling you that you are a high school teacher, or an AI specialized in reviewing accident reporting forms. Along with the role of AI, it is also recommended describe the task to be performed. Come on, this is where you have to tell him what you want him to do by playing the role that you have indicated. These two elements should be the beginning of a good prompt. 2. Add dynamic content to contextualize After describing the task, it is helpful to provide content to better contextualize your task or the content you want me to work on. For example, if you want it to analyze a text, a photograph, or even a web link, you should add it after the description. This content can also be another element that you have obtained from an application, or even from an AI, whether it is a screenshot or something else. The main thing is that add what the AI ​​needs to do what you asked it to do. You can attach several files, but then you will have to describe well what you want it to do with each one. 3. Detailed task instructions At the beginning of the prompt you have told it in summary what you want it to do, and then you have attached the content on which you want it to work. Now, after this introduction you will have to detail task instructions What do you want me to do? This fragment of text becomes the heart of the task you are programming. You will have to tell it what you want it to do precisely, just as if you were telling it to a person so that they understand it correctly without having to ask you anything else. If you have added several images or different types of content for context, you will also have to explain what you want it to do with each of the elements. 4. You can give him examples of what you want him to do When you do not give the AI ​​any example of the result you want it to offer you, it is what is called a Zero-shot or “0 examples”. With this, you will blindly trust that the artificial intelligence model knows how to do the task. However, when you want to obtain very specific results or with a specific formatthen you should explain this with examples. You can use a single example or multiple examples. By doing this, the model will see the examples of the task already classified or solved, and will use that information to generate the new answers. keeping the same response format that you have indicated. The more examples you give, the more precise the answer format will be. This is optional, but in tasks that require a specific type of response it can be very useful. 5. Repeat critical instructions. You’ve started by describing the task, and then you’ve given him a thorough description of each step he must take. But if you’re asking it to do a very complex task with a particularly long prompt, it’s a good idea to at the end repeat the most critical instructions of the task you have asked him to do. This is the equivalent of underlining the most important and vital points of the instructions, something that you think is absolutely vital that you always keep in mind and not overlook it. Advanced prompt structure in 10 steps If the five-step structure isn’t enough for you, Claude’s creators also aim for a 10-step structure. It is like the first one we told you about, but more fragmented to give AI each and every detail what you will need. Each of these steps can be one or several paragraphs in the prompt that we are going to compose. You don’t have to do this every time, normally the previous structure is enough, but for particularly complex tasks can help you. … Read more

Log In

Forgot password?

Forgot password?

Enter your account data and we will send you a link to reset your password.

Your password reset link appears to be invalid or expired.

Log in

Privacy Policy

Add to Collection

No Collections

Here you'll find all collections you've created before.