跪拜 Guibai
← All articles
Frontend · DeepSeek · AI Programming

DeepSeek Doubles API Prices During Peak Hours as the Subsidy Era Winds Down

By kyriewen ·
Read original on juejin.cn ↗ Google Translate ↗ Alt translation

The era of subsidized, near-free API access to frontier models is ending just as production reliance on them deepens. Budgets that assumed flat per-token costs now face a 2x multiplier during the hours that actually matter, forcing a hard conversation about cost engineering, workload scheduling, and vendor lock-in.

Summary

DeepSeek notified users that its V4 Pro and V4 Flash model APIs will charge double the standard rate during peak hours. The change directly raises production costs for services that depend on the platform, since most business traffic naturally clusters in those windows. Teams are being advised to recalculate their bills and consider shifting workloads to off-peak periods.

The price adjustment lands alongside a State Council meeting that prioritized domestic AI infrastructure and an "AI+" industry-wide mandate, signaling top-level support for homegrown models and chips. Meanwhile, South Korea's Samsung and SK Group committed a combined 4,755 trillion won to semiconductors, and the EU Council finalized a simplified AI regulation that pushes high-risk system compliance deadlines to late 2027.

Ford's experience offers a counterpoint to the AI-everywhere narrative: after large-model quality control systems amplified design flaws, the automaker rehired 350 veteran engineers and subsequently topped J.D. Power's 2026 initial quality rankings for the first time in 16 years.

Takeaways
DeepSeek V4 Pro and V4 Flash API prices double during peak hours while off-peak rates stay unchanged.
Most production traffic hits during peak windows, so the effective cost increase for many teams will approach 2x.
China's State Council is pushing accelerated construction of ultra-large-scale computing clusters and an "AI+" mandate across all industries.
South Korea's Samsung and SK Group plan a combined 4,755 trillion won investment in semiconductors and physical AI, with DRAM capacity expected to double in five years.
The EU Council finalized simplified AI rules, delaying high-risk system obligations to end of 2027 and watermarking requirements to December 2026.
Ford scrapped a large-model quality-control system after it amplified design defects and rehired 350 veteran engineers, then won J.D. Power's top quality ranking for the first time in 16 years.
Conclusions

DeepSeek's price hike follows a classic platform playbook: subsidize adoption, build dependency, then monetize the peak-demand hours that customers cannot easily shift.

Policy signals from Beijing and Seoul are converging on the same bet—that domestic control over compute infrastructure and memory supply chains is a strategic necessity, not just an industrial one.

The EU's delayed compliance deadlines buy time but don't change the regulatory trajectory; companies that treat the extension as a pause rather than prep time will face a scramble later.

Ford's reversal is a concrete data point against the assumption that expert tacit knowledge can be compressed into training data. The model didn't just fail to replace engineers—it actively made things worse.

The simultaneous arrival of API price normalization, infrastructure nationalism, and a high-profile LLM-in-production failure marks a shift from unbounded AI optimism to cost-and-capability accounting.

Concepts & terms
Peak-hour API pricing
A pricing model where API calls cost more during high-traffic periods, reflecting real infrastructure load and letting providers capture revenue from inelastic demand.
AI Act Omnibus amendment
A set of revisions to the EU AI Act that streamline certain obligations, including postponing compliance deadlines for high-risk AI systems and transparency requirements.
Physical AI
AI systems that interact with the physical world, such as robotics and autonomous manufacturing, as distinct from purely digital or software-based AI applications.
Source: juejin.cn ↗ Google Translate ↗ Backup ↗