DeepSeek could increase prices by 30 times while remaining the most cost-effective solution thanks to its cache architecture

DeepSeek announced on August 6th plans to implement a substantial price increase across all its API services; details of the new pricing structure will be provided in an official notification. This news emerged just weeks after the V4 Flash model ranked first in the weekly global token usage leaderboard on OpenRouter, achieving 7.22 trillion tokens.

The main offerings remain V4 Pro and V4 Flash. The current price for V4 Flash is 0.02 yuan per million tokens for cached inputs, 1 yuan for uncached inputs, and 2 yuan for output.

The price increase is attributed to rising computational costs. Previously, in April, DeepSeek had lowered the price for V4 Pro cached inputs and ran a limited promotion offering a 75 percent discount until May 5th. On May 23rd, the company announced that after the promotion ended on May 31st, the price would rise to one-quarter of the original retail rate.

In the same week, Xiaomi's MiMo-V2.5 API announced a permanent price reduction of up to 99 percent, leading to a sharp collapse in the price war within the domestic market.

A turning point occurred at the end of June. In an email from DeepSeek, users were informed that the official V4 version was scheduled for mid-July, and this new release would include pricing based on peak and off-peak hours. During peak weekday load times, API prices will double. For V4 Pro, the price for cached inputs will increase from 0.025 to 0.05 yuan per million tokens, for uncached inputs—from 3 to 6 yuan, and for output—from 6 to 12 yuan. V4 Flash will also double its tariffs in a similar manner.

This current round of increases is a general uplift. According to industry analysis, the previously introduced structure with peak and off-peak hours was more of a standardization tool for distributing scarce computing resources than a direct price hike, as it involved predictable timeframes and multipliers. The new phase raises the baseline tariff level. DeepSeek's computation costs have grown in sync with demand growth. For instance, on August 1st, the OpenCode platform registered 8 trillion tokens of V4 Flash traffic, and on August 4th, the V4 Flash API experienced overload due to unprecedented incoming traffic volume.

DeepSeek is not alone in these trends. At a briefing on Zhipu AI's performance, CEO Zhang Peng announced that in the first quarter of 2026, API prices had increased by 83 percent compared to the previous year, while API call volume grew by 400 percent. In June, Doubao Pro introduced a three-tiered subscription priced at 68, 200, and 500 yuan per month. In July, Kimi suspended new consumer registrations after K3 launch traffic exceeded cluster capacity for 48 hours. The entire industry is shifting from a 'lowest price wins' strategy to a 'sustainable cost structure' model. Training and maintaining advanced models on Chinese computing power are no longer a loss leader endlessly subsidized by the rest of the stack.

The DeepSeek V4 Flash model demonstrated massive demand, consuming 8 trillion tokens in a single day on the OpenCode tool. This volume exceeded the average daily usage of the OpenRouter platform, which is around 6.6 trillion tokens. This surge in usage reflects a fundamental shift in AI interaction methods: where tasks used to be simple question-and-answer exchanges, agents now perform more complex operations such as reading files, modifying code, running programs, and retrying after errors, which significantly increases token consumption.

An example of such complex interaction was shown by a user who created a single-page 3D game project using Claude Opus 5. This process resulted in one HTML file that required 690 million tokens and cost nearly 3000 yuan.

With the advent of agent-driven tasks, the evaluation criterion for models has shifted from the quality of a single response to the cost of completing a long-running task. For instance, one million output tokens cost 2 yuan with DeepSeek, whereas the same operation with Claude Opus 4.8 cost $25 USD, equivalent to approximately 170 yuan, demonstrating an 85-fold price difference. Although the V4 Flash Max model scored 50 points compared to 56 for Opus Max in the Artificial Analysis Intelligence Index v4.1, the cost to pass this test was $72.02 for V4 Flash versus $3752.55 for Opus, with Opus gaining 6 extra points at roughly a 52-fold increase in call costs.

V4 Flash cannot yet prove absolute superiority, but if both models appear in the same candidate list, the 85-fold price difference could change the standard selection. When a model only slightly outperforms its competitor but costs dozens of times more, the mathematics of choice no longer favors the more expensive option.

The first to feel the impact of this trend are not the weakest models, as smaller models remain suitable for routing, and flagship models continue to handle the most complex tasks. Mid-tier models are the most vulnerable—they are stronger than V4 Flash but cost dozens of times more and are not powerful enough to solve the tasks that V4 Flash handles. This is DeepSeek's 'killer line': models do not necessarily need to be the strongest; they just need to be good enough to perform most tasks, leveraging a price gap of almost two orders of magnitude to displace more expensive options from the default list.

In terms of pricing, Opus costs $25 per million output tokens, Fable 5 costs $50, and Chinese models include Qwen 3.8 Max priced at 36 yuan, Kimi K3 at 100 yuan, and GLM 5.2 at 28 yuan.

The ability to sell Mutai at mineral water prices is due to the direct inference cost per token, which depends on activated parameters, accuracy, attention, KV cache, output length, hardware utilization, and parallelism. V4 Flash has 284 billion total parameters but activates about 13 billion per token by using hybrid attention, low-precision weights, and cache optimization. With a context of one million tokens, the inference computational load is about 10% of V3.2, and the KV cache size is about 7%. The V4 series overhauled the attention mechanism, implementing CSA compress-then-focus and HCA full-browse mechanisms, and utilizes DeepSeekMoE routing, directing each token to a small subgroup of experts. Commercial pricing and service costs follow these technical features. Thus, the 'killer line' redefines the boundary of development: the combination of performance and price becomes the decisive factor, where the cost of task execution is the determining metric.