DeepSeek Warns of Major Price Hike for AI Models After Dominating with Rock-Bottom Rates

DeepSeek is preparing users for a big price hike

Notably, this is separate from a mid-July price increase announced by the company for peak-hour usage to ease server loads. Among AI providers like OpenAI, Google, and Anthropic, China-based DeepSeek has emerged as a more economical option for many. The company earned some positive headlines recently following the launch of its powerful V4 Flash model. Well, there's some bad news coming today, as DeepSeek is seemingly planning a price hike. According to a report by Bloomberg, DeepSeek is sending notices to its clients, informing them of a "significant" price increase. There's no word on how much more expensive its offerings will be, nor on when the updated prices will take effect. The general idea behind the notice was to give users enough time to prepare for the imminent changes. For some background, DeepSeek currently charges rates of $0.28 per million output tokens and $0.14 per million input tokens for the V4 Flash model. By comparison, Google's cost-effective Gemini 2.5 Flash-Lite model costs $0.10 per million input tokens and $0.40 per million output tokens. However, the actual cost depends on each individual's usage, as some may require heavy input tokens, while others may consume more output tokens. This upcoming price increase is separate from the price hikes announced for peak-hour operation last month. This meant DeepSeek would effectively double API usage rates during certain time slots of the day to ease pressure on its servers. As another Bloomberg report pointed out last week, DeepSeek has set out plans to build a massive data center in Inner Mongolia to bolster its AI infrastructure. This upcoming wave of price hikes could be one of the multiple ways the company intends to fund the data center, though it is worth noting that DeepSeek hasn't officially confirmed this project.

DeepSeek warns of a 'significant' price rise, reversing its cheap-AI pitch

The Chinese lab that made cheap models its calling card has warned of a significant API price rise, a telling reversal for the company that started the race to the bottom. The company that made AI cheap is about to make it less so. DeepSeek has warned that a significant price increase is coming for its API, a striking about-face for a firm built on undercutting everyone else. The details are still thin. DeepSeek said only that the rise would land in the near future and would apply across its API pricing, without publishing a new schedule or naming figures. What makes it notable is who is doing it. DeepSeek is the lab whose rock-bottom prices turned the AI market into a price war, forcing rivals to slash their own rates to compete. The timing is pointed. The warning comes barely a week after DeepSeek launched V4-Flash, a lightweight model it had been promoting as one of the cheapest capable options available anywhere. That model was a genuine hit. Independent testing had crowned V4-Flash the cheapest well-known model to run, exactly the reputation the company is now complicating with talk of higher prices. The most likely explanation is demand. DeepSeek's low prices drew a surge of users, and serving all of them strains the compute behind the service, making ultra-cheap pricing harder to sustain. The economics of cheap AI are unforgiving. Every query costs real money in chips and power, and a provider that prices below cost to win share eventually has to reckon with the bill. DeepSeek has been experimenting with the answer. It recently introduced peak-hour surge pricing, charging more when demand spikes, and a plan to double rates at busy times has been floated but not yet switched on. The competitive ground has also shifted. Cheap, capable models are no longer a Chinese speciality, and Western rivals have closed much of the gap that once made DeepSeek stand out. One developer captured the mood. Michael Guo questioned the timing, noting that Meta's Muse Spark and OpenAI's newest model now match DeepSeek on capability and price, eroding its core advantage. That is the real risk in raising prices. DeepSeek's brand is cost-effective AI, and if it charges more just as rivals get cheaper, it hands them the very differentiator it spent a year defending. The move also carries a message about the whole sector. If even the cheapest provider must raise prices, the era of AI subsidised to near-free for users may be reaching its limits. DeepSeek's rise was built on that subsidy. Its models stunned the industry by matching Western rivals at a fraction of the cost, a shock that wiped value off chipmakers and forced a global rethink of how much frontier AI should cost. Raising prices tests how much of that was real. If demand holds even as rates climb, it proves DeepSeek built genuine loyalty; if users flee to cheaper options, the cost advantage was the whole business. There is a strategic reading too. Higher prices could steady DeepSeek's finances and signal a shift from land-grab to sustainability, the same maturation every disruptive challenger eventually faces once the growth is banked. For customers who built on DeepSeek precisely because it was cheap, the warning is a prompt to plan. Higher costs could push some toward rivals, testing how much loyalty low prices actually bought. Rivals will be watching closely for an opening. Every cent DeepSeek adds is a cent a competitor can undercut, and the labs that spent a year chasing its prices will happily pitch themselves as the new cheap option. For now, the increase is a signal rather than a number. But coming from the company that set the pace of the price war, even a warning is enough to make the whole market recalculate.

DeepSeek to get a 'significant' price hike soon

Everything is becoming more expensive, including things that are supposed to be cheap. The latest example is DeepSeek's AI services, as reported by Bloomberg. The Chinese AI firm sent a notice to users on Thursday warning of a "significant" price increase, though it did not specify just how significant it would be. DeepSeek has been known for its extremely cheap services relative to the competition, especially with regards to its western counterparts, so this is probably not the most welcome news among DeepSeek users. To be more specific, DeepSeek currently charges less than a dollar per million input and output tokens, which is how AI services are sold to customers. By comparison, Anthropic's Fable 5 model charges $10 per million input and $50 per million output tokens. It was probably unsustainable for DeepSeek to continue charging so little compared to the competition, especially as the company plans to build a huge data center in Inner Mongolia. But still, you can't count on anything staying cheap anymore. Not even AI, the thing that's making everything else expensive in the first place.

China's DeepSeek Upgrades V4 Pro: Claude Fable Is Only 5% Better at 4,500% the Price

Fable 5 costs $10 per million input tokens and $50 per million output. V4 Pro costs $0.435 and $0.87. DeepSeek shipped the finished version of its flagship on Wednesday with no blog post and no announcement. The tell was a table cell: the model version listed for "deepseek-v4-pro" on the API pricing page now reads DeepSeek-V4-Pro-0813. The cost to use this model is around $0.435 and $0.87 per million tokens (the basic unit of information a model can handle) of input and output. So the pricing structure remains the same, but what changed is the weights underneath. Deepseek V4 Pro has been in the wild since April, priced 98% below GPT-5 Pro, and every independent lab that tested it was testing a preview. DeepSeek said so itself on July 31, when it pushed V4-Flash to general availability and noted that the Pro API was "unchanged" with the official release to "follow soon." The model card on Hugging Face, the go-to repository for open-source AI projects, still describes the V4 series as "a preview version." So the widely circulated scores describe a build DeepSeek considered unfinished. Nobody outside the company has independently benchmarked 0813 yet. What DeepSeek's table claims The company published a comparison across 10 agent benchmarks. On the eight where Fable 5 or another model has the advantage, the gaps are small. Average Fable 5's relative lead across the benchmarks and you get 5.3%. Strip out Humanity's Last Exam without tools -- where DeepSeek scores 42.7 against 53.3, a 10.6% gap that skews everything -- and the remaining rows average 2.8%. Pricing is public on both sides, and this is where the comparison stops being close. Fable 5 runs $10 per million input tokens and $50 per million output. V4 Pro runs $0.435 and $0.87, with cached input at $0.003625. On blended rates that's $30 against $0.65 -- roughly 46 times, or 4,600% of the cost. That's the kind of spread that matters when you're running a business and using AI tools at scale. Cost per completed task runs wider, because Fable 5 thinks longer and writes more. Artificial Analysis measured it at $3.15 per benchmark task against 3 cents for V4-Flash, about 105 times cheaper. Hugging Face CEO Clément Delangue put the spread at over $31 per task against roughly $0.04. No per-task figure exists for 0813 yet. Anthropic's own lineup complicates the premium. Claude Opus 5 outscores Fable 5 on most benchmarks at half the price. DeepSeek scored these itself, on infrastructure it hasn't released. Its July note specified DeepSeek Harness minimal mode "to be released soon," running at max effort with high creativity. Two of the ten benchmarks, DSBench-FullStack and DSBench-Hard, are internal test sets with no public leaderboard to check them against. The direction of travel is consistent, though. Chinese open-weights labs keep landing within a few points of the American frontier at a fraction of the price -- Kimi K3 beat Fable 5 and GPT-5.6 Sol on release, and DeepSeek and Xiaomi have been cutting frontier costs by 99% while U.S. labs go the other way. DeepSeek's weights are MIT-licensed and on Hugging Face, so independent verification is a download away.

DeepSeek Prices Its New V4-Pro-0813 Model At $0.87 Per 1 Million Output Tokens, As The High-Flying Chinese AI Lab Wows With Its Soaring Token Consumption

DeepSeek was second only to Anthropic in terms of the total number of tokens consumed in July. And now, perhaps in a bid to cement its ascendancy, the high-flying Chinese AI lab has just unveiled the DeepSeek-V4-Pro-0813 model, its latest gambit to take on the might of OpenAI and Anthropic. The V4-Pro-0813 is now rolling out on DeepSeek API and DeepSeek Chat DeepSeek has started rolling out the V4-Pro on its API and Chat. The AI lab has priced the model at $0.435 per 1 million tokens of input and $0.87 per 1 million tokens of output. Do note that OpenAI launched a literal price war a few days back by discounting its GPT-5.6 Luna by as much as 80 percent, with input tokens now priced at just $0.2 per 1 million from their earlier perch at $1, and output tokens priced at just $1.20 per 1 million vs. the earlier price of $6. Just hours later, however, DeepSeek launched a refreshed version of its latest Flash-class model, dubbed the V4-Flash-0731. Critically, the model has just 284 billion parameters and yet offers a performance that is similar to Anthropic's Opus 4.8, which is widely believed to span multi-trillion parameters! And, in what went right to the heart of OpenAI's price war, DeepSeek priced the V4 Flash 0731 at just $0.14 per 1 million tokens of input, and $0.28 per 1 million tokens of output, eviscerating any comparative price advantage that OpenAI tried to garner with its discounting move. Coming back, DeepSeek's V4-Pro model outcompetes Opus 4.8 on Terminal Bench 2.1, Cybergym, DeepSWE, and AutomationBench benchmarks, as per the preliminary results populating WeChat right now. Meanwhile, as stated earlier, DeepSeek was second only to Anthropic in terms of token volume in July, and might even clinch the apex spot in the coming months. As such, DeepSeek is currently contending with an unprecedented demand surge, especially amid anecdotes that suggest its models' inference speeds slow down to a crawl at times, which is wholly understandable given the lab's limited compute footprint of just around 20,000 NVIDIA H100 GPUs. Follow Wccftech on Google to get more of our news coverage in your feeds.

DeepSeek Forced To Raise Prices As Its Recent Price Cuts To Snub OpenAI Unleashed A Demand Tsnunami That Its 20,000-GPU Stash Can't Handle

Too much of a good thing can become poison, as DeepSeek appears to be finding out now after it nipped OpenAI's nascent price war in the bud by aggressively cutting the price of its V4 Flash model, unleashing a demand tsunami that its infrastructure simply can't keep up with. DeepSeek unleashed a tsunami of demand with its recent price cuts for the V4 Flash model, forcing an abrupt about-face As we detailed recently, OpenAI launched a literal price war a few days back by discounting its GPT-5.6 Luna by as much as 80 percent, with input tokens now priced at just $0.2 per 1 million from their earlier perch at $1, and output tokens priced at just $1.20 per 1 million vs. the earlier price of $6. Even though OpenAI claimed at the time that it was able to implement this steep discount after extracting additional architectural efficiencies from its models, most interpreted the move as its opening gambit in a price war aimed at China's AI labs. Just hours later, however, DeepSeek launched a refreshed version of its latest Flash-class model, dubbed the V4 Flash 0731. Critically, the model has just 284 billion parameters and yet offers a performance that is similar to Anthropic's Opus 4.8, which is widely believed to span multi-trillion parameters! And, in what went right to the heart of OpenAI's price war, DeepSeek priced the V4 Flash 0731 at just $0.14 per 1 million tokens of input, and $0.28 per 1 million tokens of output, eviscerating any comparative price advantage that OpenAI tried to garner with its discounting move. Today, however, DeepSeek has started informing its customers that its API prices are all set for a "significant increase." While the AI lab has given no reason for this aberrant move - which would only affect API pricing and not customers who run these open-weight models on their own infrastructure - the word on the street is that the company is currently contending with an unprecedented demand surge, especially amid anecdotes that suggest its models' inference speeds slow down to a crawl at times, which is wholly understandable given the lab's limited compute footprint of just around 20,000 NVIDIA H100 GPUs. Perhaps DeepSeek has bitten off more than what it can chew at this time, giving OpenAI a much-needed win in its nascent price war. Follow Wccftech on Google to get more of our news coverage in your feeds.

DeepSeek warns of 'significant' price hike for its API services By Investing.com

Investing.com-- Chinese artificial intelligence startup DeepSeek said it plans to significantly increase prices for its application programming interface (API) services as demand for its latest AI models accelerates. The company said on its developer documentation page that it intends to raise overall API pricing "in the near future," with a "significant increase expected," although it did not disclose the timing or the size of the increase. Get real-time updates on market-moving news with InvestingPro -- at 55% off now It urged customers to plan their usage accordingly, adding that the final pricing structure would be announced in a separate official notice. The planned increase comes days after DeepSeek launched its latest models, DeepSeek-V4-Flash-0731 and DeepSeek-V4-Pro, which support up to a one million-token context window and expanded reasoning capabilities. Under the current pricing, DeepSeek charges as little as $0.14 per one million input tokens for cache misses on its V4-Flash model and $0.28 per one million output tokens, while its V4-Pro model costs $0.435 and $0.87, respectively. Research firm Artificial Analysis said DeepSeek's V4-Flash currently costs about $0.03 per benchmark task, compared with roughly $0.86 for Moonshot AI's Kimi K3, $1.86 for OpenAI's GPT-5.6 Sol and $3.15 for Anthropic's Claude Fable 5, underscoring the Chinese firm's pricing advantage. The move could narrow DeepSeek's pricing advantage over rivals as competition in the AI model market intensifies.