DeepSeek V4 Pro Price Cut: Chinese AI Startup Permanently Lowers Flagship Model API Rates by 75% to Challenge Global Competitors
Chinese artificial intelligence startup DeepSeek has made its 75% promotional discount on the flagship V4 Pro model permanent. Optimized for Huawei's Ascend 950 chips, the API pricing now scales down to CNY 0.025 per million tokens. This aggressive cost reduction creates an intense pricing war, heavily undercutting Western competitors like OpenAI.
Chinese artificial intelligence startup DeepSeek has announced that its highly publicised 75 per cent discount on its flagship large language model, V4 Pro, will become its permanent retail pricing floor. The promotion, which was originally scheduled to expire on May 31, 2026, repositions the Hangzhou-based firm as a long-term cost leader in the global generative software industry. By locking in these aggressive rates, the startup is directly challenging its Western foundation model competitors, including OpenAI and Anthropic, by pricing its frontier-level developer access significantly below standard market expectations.
DeepSeek V4 Pro: API Costs Adjusted Across Token Metrics
Following the structural price revision, DeepSeek’s V4 Pro application programming interface (API) costs have been set to a permanent range of between CNY 0.025 and CNY 6 per million tokens, or approximately USD 0.0035 to USD 0.83 depending on specific computational tasks. This is a dramatic decrease from the model’s initial launch pricing tier, which demanded between CNY 0.1 and CNY 24 for the equivalent volume of processed data. A token represents the core unit of text processed by an artificial intelligence architecture during user interactions. DeepSeek V4 Pro, DeepSeek AI model, DeepSeek API pricing, Chinese AI startup, DeepSeek vs OpenAI, DeepSeek vs Anthropic, DeepSeek, OpenAI, Anthropic, AI Model
DeepSeek V4 Pro Available at Massive Discount
We are making our discount permanent! 🎉
Enjoy building with DeepSeek-V4-Pro and bring your innovative ideas to life! 🚀 https://t.co/ctbZely7of pic.twitter.com/V8atbTaogH
— DeepSeek (@deepseek_ai) May 22, 2026
The move signals a deliberate strategy by the Chinese developer to capture extensive enterprise market share through structural cost advantages rather than temporary client acquisition tactics. By transforming a brief promotional discount into a fixed baseline price, the company has established a remarkably low fiscal ceiling that international technology firms may struggle to match without suffering notable hits to their operating profit margins.
Huawei Semiconductor Infrastructure Realignment
The permanent price reduction highlights an expanding relationship between domestic artificial intelligence software and homegrown hardware infrastructure within the region. DeepSeek’s V4 Pro system was deliberately built and optimized to run on Huawei Technologies' Ascend 950 series processors rather than American silicon. When the 1.6-trillion-parameter model originally launched last month, DeepSeek noted that the Pro version initially cost 12 times more than its Flash equivalent due to acute constraints in high-end computing capacity.
While DeepSeek did not explicitly clarify whether an immediate increase in Huawei's semiconductor supply triggered the permanent price drop, it previously stated that operational costs would fall sharply as Ascend 950 supernodes deployed at a mass scale. Huawei's advanced hardware sales have expanded substantially in recent quarters, capitalizing on strict United States export controls that prevent competitors like Nvidia from distributing their most advanced hardware architectures into the Chinese domestic market.
Geopolitical Dynamics in Contemporary Machine Learning
The price war unfolds against a backdrop of escalating technical friction between Washington and Beijing over international semiconductor supply chains. While strict trade curbs on high-end chipmaking equipment continue to limit Huawei’s capacity to seamlessly ramp up its total manufacturing output, domestic demand for its Ascend chips has surged as regional internet giants look to secure their long-term infrastructure. OpenAI GPT-Realtime-2 Launched With Advanced Reasoning and Voice-to-Action Capabilities; Check Features.
By delivering a high-capacity model that operates independently of Western hardware at a fraction of the cost of its American counterparts, DeepSeek is accelerating the development of a parallel technology ecosystem. As local cloud providers deploy the V4 Pro API across their corporate platforms at these newly locked-in rates, the broader artificial intelligence industry faces an asymmetric pricing landscape where structural operational costs have become the definitive competitive battlefield.
(The above story first appeared on LatestLY on May 23, 2026 11:34 PM IST. For more news and updates on politics, world, sports, entertainment and lifestyle, log on to our website latestly.com).
