AI
May 23, 2026DeepSeek Makes the V3 Pro Price Discount Permanent
DeepSeek has locked in the discounted pricing for its V3 Pro model, removing the temporary label from what was previously a promotional rate.
DeepSeek has made the discounted pricing on its V3 Pro API permanent. The rate, previously offered as a limited-time promotion, is now the standard price listed in the official pricing documentation.
This matters for builders running cost projections. Token budgets modeled around the promotional rate no longer carry expiry risk. Teams that deferred committing to V3 Pro because the discount had an implicit end date can now build on a stable cost floor.
DeepSeek's API pricing has been aggressive relative to comparable frontier models. V3 Pro sits in the lineup above the base V3 tier and targets workloads that need higher throughput or quality headroom without jumping to the R1 reasoning model, which carries a different cost structure. Locking in the discounted rate at the lower tier keeps the price gap between the two meaningful.
For solo founders and small teams, the calculus shifts slightly. Permanent pricing means infrastructure cost estimates survive into production without renegotiation risk. It also signals that DeepSeek is treating this tier as a long-term product rather than a trial offering.
The announcement does not introduce new capabilities or change rate limits. It is strictly a pricing policy update. Engineers already running on V3 Pro see no immediate action required — the benefit is forward-looking: budget certainty and reduced procurement friction for new projects.
DeepSeek has been publishing its API pricing and model details at api-docs.deepseek.com. The pricing page reflects the change directly, with no separate announcement post. Teams evaluating the model for production use should review the current tier breakdown there to confirm input, output, and cache pricing against their specific token distribution before committing.
Source
news.ycombinator.com