Latest news with #EricSimons
Business Insider
2 days ago
- Business
- Business Insider
As AI coding services face a reckoning, Bolt tries to go beyond building
AI coding startups are facing a reckoning, with questions about pricing models and customer churn. Eric Simons, CEO of startup StackBlitz, has been working on potential solutions. StackBlitz runs a popular AI coding service called and on Thursday the startup rolled out new features and an updated subscription aimed at keeping customers engaged on its platform for longer, Business Insider learned exclusively. In the crowded AI coding market, many startups rely on reselling AI inference. They pay for access to AI models from frontier labs and Big Tech providers, then charge customers for their bespoke coding services — hoping to make a profit on the difference. However, once users finish building a project, there's less reason to stick around. That increases the chance that some customers cancel their subscriptions. Simons is seeing churn rates across the AI coding market running at 20% to 40%. He declined to say what Bolt's churn rate is; however, he noted that Wix, a more traditional website-building service, has much lower churn rates. That's because Wix generates the bulk of its recurring revenue from hosting and similar stickier offerings, rather than one-off project creation. "This is the problem across all these companies right now. The churn rate for everyone is really high," Simons said. "You have to build a retentive business." Love Business Insider? Log into Google and make us a preferred source. Adding more valuable, stickier services on top of these models is the goal now, and Simons is repositioning Bolt with a new mantra, "Build Without Boundaries." "The challenge here is that if your business is purely AI, reselling AI inference, your business is very fragile and vulnerable, because the winds can shift violently," Simons added. One solution is to offer a broader array of tools that bring more users into your ecosystem. This is why Bolt launched a subscription overhaul on Thursday aimed at making the service more of an end-to-end platform, not just a building tool. Pricing will remain the same for most tiers, with only the entry-level plan rising from $20 to $25 a month. Every plan will now include hosting, domains, databases, serverless functions, authentication, SEO optimization, Stripe-powered payments, and analytics for an unlimited number of projects. Bolt is partnering with companies such as infrastructure leader Netlify and database provider Supabase to provide the same scale and reliability used by major tech companies, but wrapped into one subscription. The idea is to keep users in the Bolt ecosystem well past the initial creation phase. Instead of exporting their projects to another host or backend service, users can launch, run, and scale them without leaving the Bolt platform. Pay-as-you-go pricing will kick in only for unusually high traffic, with user-set caps to avoid surprise bills, Simons told Business Insider. Simons sees this approach resonating with two main audiences: "prosumers," solo builders and entrepreneurs launching new products; and B2B product teams that use Bolt for rapid prototyping and internal tools. For the latter, staying in Bolt isn't just convenient; it embeds the platform into their ongoing workflow, making it harder to leave. By going "beyond building" and integrating the entire product lifecycle into one subscription, Simons hopes to shift Bolt from being a one-off AI tool to an essential part of its customers' day-to-day operations. Indeed, other AI coding services, Replit and Lovable, offering extra features now too, such as hosting. In an industry where the winds can change fast, the bet is that offering a complete, durable ecosystem is the only way to scale sustainably.
Business Insider
4 days ago
- Business
- Business Insider
'Inference whales' are eating into AI coding startups' business model
The AI coding sector has a problem. Heavy users of AI coding services have been racking up huge costs, forcing some leading startups to overhaul their pricing structures and offerings to avoid big losses. "Inference whales," as some in the business call these customers, are making industry insiders question whether AI products that are just "reselling inference" can survive long-term. Inference refers to how AI models are run. Newer reasoning models break user requests down into multiple steps, which increases inference costs. When applied to AI coding services, where developers set automated agents to longer-term tasks, expenses can soar quickly. This is a problem for AI coding services because they're often offered through monthly subscription plans. Many plans allow unlimited use for a fixed monthly fee, and a few users have taken advantage by bombarding the services with huge projects. These startups must still pay for the underlying AI models, so they're getting squeezed between a relatively fixed revenue stream and rapidly rising backend costs. "If you're purely reselling AI inference, your business could be very fragile and vulnerable, because the winds can shift violently," said Eric Simons, CEO of StackBlitz, and startup that offers a popular AI coding service called Claude Code whales Anthropic offered its popular Claude Code service through a $200 a month unlimited plan earlier this year. Some subscribers went berserk, using thousands of dollars' worth of AI inference over a few weeks or months. Someone even built a website to rank these AI coding whales. The Claude Code Leaderboard lists one developer at the top who's burned through almost 11 billion tokens. Tokens are how AI models break queries down into digestible data chunks. Industry pricing is based on how many tokens are processed. That top-ranked developer's token usage costs almost $35,000, according to this leaderboard. That compares to the $200 a month he's been charged. Even if that's over a whole year, Anthropic would be getting about $2,400, while incurring much higher inference costs. Anthropic is changing its pricing That's clearly unsustainable, so Anthropic plans to change its pricing. The $200 a month plan will stay, but the startup will introduce weekly rate limits, starting August 28. If users blow through these new weekly rate limits, they will have to buy additional capacity. "We've identified extreme usage by a small number of customers that impacts capacity for our broader community," an Anthropic spokesperson told Business Insider. The startup said it's also seen "policy violations," such as account sharing and reselling access. "We're committed to supporting advanced use cases long-term, but need to ensure consistent performance for all developers in the meantime," the Anthropic spokesperson added. A Swedish whale I tracked down one of the whales near the top of the Claude Code Leaderboard. Albert Örwall, a developer based in Sweden, said he's been using the $200 a month Claude Code subscription to build his own vibe-coding platform, along with some open-source agentic tools. "I was probably running 3 to 4 fairly long-running tasks in parallel constantly while I was working, and that's when it really took off," he said of his Claude Code usage. Even excluding these big projects, Örwall said his regular workflow in Claude Code likely racks up inference costs of $500 per day, under a subscription that costs only $200 a month. "So I'm guessing my workflow might not be sustainable for Anthropic," he added. Cursor responded, too When Anthropic's new pricing kicks in, Örwall said he'll keep the $200 a month subscription for a while to get a feel for what the weekly limits actually mean for his budget. "I'll avoid paying anything beyond the $200 subscription," he said, noting that he can change how he writes code and develops projects to avoid breaching the new rate limits. "The reason I originally switched from Cursor to Claude Code was because usage-based pricing became too expensive in Cursor," Örwall added. Cursor is another popular AI coding service, which often uses Anthropic's AI models as the underlying intelligence powering its product. Cursor recently switched its $20 a month Pro plan from unlimited requests to a tiered system with usage-based pricing for "fast" requests, meaning users are charged extra for exceeding a certain limit. This change, coupled with a lack of clear communication, caused confusion and frustration among some users who expected unlimited usage. Cursor announced the initial change in mid-June. Then it updated with more details about 2 weeks later, and then again in early July. "New models can spend more tokens per request on longer-horizon tasks," the startup wrote in a blog post, apologizing for surprising users with unexpected new bills. "Though most users' costs have stayed fairly constant, the hardest requests cost an order of magnitude more than simple ones." Inference costs aren't falling The assumption across the industry has been that inference costs will drop dramatically, making these AI coding services more financially viable. However, in practice, this hasn't happened thus far. Instead, when a new top AI model comes out, all the AI coding services integrate it — along with its higher prices. "This is the first faulty pillar of the 'costs will drop' strategy," Ethan Ding, CEO of startup TextQL, wrote in a recent blog. "Demand exists for 'the best language model,' period. And the best model always costs about the same, because that's what the edge of inference costs today." Developers and other AI users usually want the best, not last month's leading intelligence. "Nobody opens Claude and thinks, 'you know what? let me use the shitty version to save my boss some money.' We're cognitively greedy creatures," Ding wrote. "We want the best brain we can get." Even when inference costs do fall, the rise of agentic AI workflows means that developers set up longer, automated projects that generate a lot more tokens. If a project uses 100 million tokens, rather than 1 million, the initiative's cost remains high, even if per-token prices may have fallen. "A $20/month subscription cannot even support a user making a single $1 deep research run a day," Ding said. "But that's exactly what we're racing toward. Every improvement in model capability is an improvement in how much compute they can meaningfully consume." "There's no way to offer unlimited usage in this new world under any subscription model," he added. "The math has fundamentally broken."
Business Insider
4 days ago
- Business
- Business Insider
'Inference whales' are breaching AI coding startup business models
The AI coding sector has a problem. Heavy users of AI coding services have been racking up huge costs, forcing some leading startups to overhaul their pricing structures and offerings to avoid big losses. "Inference whales," as some in the business call these customers, are making industry insiders question whether AI products that are just "reselling inference" can survive long-term. Inference refers to how AI models are run. Newer reasoning models break user requests down into multiple steps, which increases inference costs. When applied to AI coding services, where developers set automated agents to longer-term tasks, expenses can soar quickly. This is a problem for AI coding services because they're often offered through monthly subscription plans. Many plans allow unlimited use for a fixed monthly fee, and a few users have taken advantage by bombarding the services with huge projects. These startups must still pay for the underlying AI models, so they're getting squeezed between a relatively fixed revenue stream and rapidly rising backend costs. "If you're purely reselling AI inference, your business could be very fragile and vulnerable, because the winds can shift violently," said Eric Simons, CEO of StackBlitz, and startup that offers a popular AI coding service called Claude Code whales Anthropic offered its popular Claude Code service through a $200 a month unlimited plan earlier this year. Some subscribers went berserk, using thousands of dollars' worth of AI inference over a few weeks or months. Someone even built a website to rank these AI coding whales. The Claude Code Leaderboard lists one developer at the top who's burned through almost 11 billion tokens. Tokens are how AI models break queries down into digestible data chunks. Industry pricing is based on how many tokens are processed. That top-ranked developer's token usage costs almost $35,000, according to this leaderboard. That compares to the $200 a month he's been charged. Even if that's over a whole year, Anthropic would be getting about $2,400, while incurring much higher inference costs. Anthropic is changing its pricing That's clearly unsustainable, so Anthropic plans to change its pricing. The $200 a month plan will stay, but the startup will introduce weekly rate limits, starting August 28. If users blow through these new weekly rate limits, they will have to buy additional capacity. "We've identified extreme usage by a small number of customers that impacts capacity for our broader community," an Anthropic spokesperson told Business Insider. The startup said it's also seen "policy violations," such as account sharing and reselling access. "We're committed to supporting advanced use cases long-term, but need to ensure consistent performance for all developers in the meantime," the Anthropic spokesperson added. A Swedish whale I tracked down one of the whales near the top of the Claude Code Leaderboard. Albert Örwall, a developer based in Sweden, said he's been using the $200 a month Claude Code subscription to build his own vibe-coding platform, along with some open-source agentic tools. "I was probably running 3 to 4 fairly long-running tasks in parallel constantly while I was working, and that's when it really took off," he said of his Claude Code usage. Even excluding these big projects, Örwall said his regular workflow in Claude Code likely racks up inference costs of $500 per day, under a subscription that costs only $200 a month. "So I'm guessing my workflow might not be sustainable for Anthropic," he added. Cursor responded, too When Anthropic's new pricing kicks in, Örwall said he'll keep the $200 a month subscription for a while to get a feel for what the weekly limits actually mean for his budget. "I'll avoid paying anything beyond the $200 subscription," he said, noting that he can change how he writes code and develops projects to avoid breaching the new rate limits. "The reason I originally switched from Cursor to Claude Code was because usage-based pricing became too expensive in Cursor," Örwall added. Cursor is another popular AI coding service, which often uses Anthropic's AI models as the underlying intelligence powering its product. Cursor recently switched its $20 a month Pro plan from unlimited requests to a tiered system with usage-based pricing for "fast" requests, meaning users are charged extra for exceeding a certain limit. This change, coupled with a lack of clear communication, caused confusion and frustration among some users who expected unlimited usage. Cursor announced the initial change in mid-June. Then it updated with more details about 2 weeks later, and then again in early July. "New models can spend more tokens per request on longer-horizon tasks," the startup wrote in a blog post, apologizing for surprising users with unexpected new bills. "Though most users' costs have stayed fairly constant, the hardest requests cost an order of magnitude more than simple ones." Inference costs aren't falling The assumption across the industry has been that inference costs will drop dramatically, making these AI coding services more financially viable. However, in practice, this hasn't happened thus far. Instead, when a new top AI model comes out, all the AI coding services integrate it — along with its higher prices. "This is the first faulty pillar of the 'costs will drop' strategy," Ethan Ding, CEO of startup TextQL, wrote in a recent blog. "Demand exists for 'the best language model,' period. And the best model always costs about the same, because that's what the edge of inference costs today." Developers and other AI users usually want the best, not last month's leading intelligence. "Nobody opens Claude and thinks, 'you know what? let me use the shitty version to save my boss some money.' We're cognitively greedy creatures," Ding wrote. "We want the best brain we can get." Even when inference costs do fall, the rise of agentic AI workflows means that developers set up longer, automated projects that generate a lot more tokens. If a project uses 100 million tokens, rather than 1 million, the initiative's cost remains high, even if per-token prices may have fallen. "A $20/month subscription cannot even support a user making a single $1 deep research run a day," Ding said. "But that's exactly what we're racing toward. Every improvement in model capability is an improvement in how much compute they can meaningfully consume." "There's no way to offer unlimited usage in this new world under any subscription model," he added. "The math has fundamentally broken."
New Indian Express
08-06-2025
- Sport
- New Indian Express
WTC Final: Bavuma's Proteas chase long-awaited global glory
A win in the World Test Championship final would not just allow South Africa to focus and invest more on red-ball cricket but also encourage the up-and-coming young talents in the country to prioritise the longer format. Former South Africa cricketer and noted coach Eric Simons, who has been on and off coaching the South African team, insists that the Bavuma led-side has nothing to lose and that they will go all out against Australia. "A lot of interest (back home) as you would expect for a team playing in a prestigious event such as the WTC. South Africa is a very passionate sporting nation and a match like this brings many people together, even those that generally do not follow cricket. The interest will no doubt pick up when the match starts and the conversations can have direct context," Simons told The New Indian Express. "Although the WTC is a relatively new ICC initiative in the context of other ICC events, it has already developed a significant status. CSA made a conscious decision to qualify for the event in spite of it being an uphill battle to reach it based on the position the team was in a few months ago. Selection of the different format squads appeared to show the importance CSA put on qualification as they made sure the best XI was available, New Zealand series being the exception for other reasons," he added. Although Australia are the defending champs and are coming into the match as favourites, Simons believes that it allows South Africa to come out all guns blazing. The fact that they have brought on Stuart Broad as a mentor to make the most of his experience with the Dukes ball could also come in handy. "Australia are justifiably the favourites but that is not a bad situation for the SA to be in. Backs to the wall and nothing to lose attitude will be a good mindset to have going into the match. It would appear they might have a slight edge, so yes, but there are a few SA players with significant experience and whose play suits the conditions. Also some local players/coaches have been drafted to assist with preparations. Also with the Duke's ball being used, both teams who play 99% of their cricket with Kookaburra will have to adapt," Simons said.
News18
06-05-2025
- Sport
- News18
Eric Simons Says CSK 'Not Going Through The Motions' After Playoffs Miss
Chennai Super Kings (CSK) bowling consultant Eric Simons clarified that his side are not 'going through the motions' after crashing out of the race for their 2025 Indian Premier League (IPL) playoffs, as they prepare to take on Kolkata Knight Riders (KKR) at the Eden Gardens on Wednesday. 'There's been some really good work done, discussions around individuals and for us as a team, obviously. We are a franchise that has a rich history of winning and doing particularly well. So it's obviously not a great place to be in, but it's been a very rewarding place for us to look at it from a perspective of growth of individuals and growth as units, so there is no going through the motions. It's still been a very rewarding period for us," Simons said at the pre-match press conference.
