From being free for use to $200 (₹17,000) a month, artificial intelligence chatbots seem to be funding their Manhattan penthouses within a short period of becoming popular.

Be it OpenAI, Anthropic, and now Perplexity, all now have a $200-per-month tier (besides other options) for subscribers. It shows individuals and companies are willing to pay for AI, which wants to become a “co-thinker” for users.

AI systems are expensive to run, especially when it comes to training, and servers are needed to keep them going, besides the best engineers in the market. The money has to come from somewhere.

Perplexity, which is the talk of Silicon Valley after Apple apparently held internal discussions to consider an acquisition (Perplexity hasn’t been approached yet), is the latest to introduce a $200 monthly plan.

People are using AI chatbots for tasks beyond just creating images of cats having coffee. From coming up with an itinerary for a vacation to planning workout routines and knowing the latest news, there is no limit to the uses of AI. The excitement on Wall Street and Silicon Valley is giving way to real-world usage.

Now, financial analysts to sports managers are using the technology. But some tasks need more AI power than others. More than 35 per cent of Americans report using some kind of AI-enabled product at least once a week, according to a recent Gallup poll. Many of these users are willing to pay.

Google is the cream on the expensive AI cake with its Google AI Ultra subscription plan that comes for $249.99 a month and is meant for the likes of filmmakers and developers. “Think of it as your VIP pass to Google Al,” said Google in a statement in May. The plan includes Flow, a new AI filmmaking tool that enables the “crafting of cinematic clips, scenes and cohesive narratives”.

The new Perplexity Max plan is for power users as it offers unlimited access to the San Francisco startup’s number crunching and report generation tools, besides early access to new features.

Like any other AI company, Anthropic, which in March closed a deal to raise $3.5 billion at a valuation of $61.5 billion, is trying to convince customers and businesses to pay for its software to help offset the costs of building AI models. The company’s latest model, Claude 3.7 Sonnet, allows users to decide if they want a quick answer to a simple question or a well-crafted response that mirrors human reasoning. The approach is giving the startup an edge over its rivals.

OpenAI was among the first to go the $200-a-month way in December. It allows access to the O1 Pro mode that wields more computing power to process answers. “Power users of ChatGPT, at this point, really use it a lot, and they want more compute than $20 can buy,” said CEO Sam Altman. The target groups are those that use generative AI models for technical work.

Paid versions of apps with AI at their heart are also becoming popular, like the AI image generator Midjourney and the design app Canva.

For content creators like Jessica Valvo from New Jersey, ChatGPT and a few other tools help her get ideas for filming videos with her bulldog named Meatball Ravioli. She is able to keep her audience engaged. Many are using AI tools to transcribe recordings, analyse spreadsheets and make PowerPoint decks.

The pricing strategy highlights a shift that’s taking place — AI from being an experimental technology to something people are depending on. At the same time, nobody knows whether this will affect how individuals spend money on a slew of subscriptions that we engage with, from streaming platforms to gaming services. After all, one’s monthly earnings remain the same.

Not just Silicon Valley, Chinese companies are also delivering several AI tools. Manus AI agent, from Butterfly Effect, is becoming popular and has a $39-per-month tier as well as a $199 upgraded option that can compete with ChatGPT Pro. The bot can undertake complex tasks rather than just responding to prompts. Some are hailing it as the next DeepSeek of the AI world.

In China, generative AI chatbots are showing growth in the mental health sector, despite concerns about the risks involved when people in distress seek help from technology.

An Elon University (it has no connection with Elon Musk) survey in the US says 38 per cent of users believe “it is very or somewhat likely that in the next decade LLMs will form deep relationships with humans”.