DeepSeek V4 Pro API Pricing: GA Rates and the Coming Price Increase

DeepSeek V4 Pro left preview on August 12, 2026. The general availability build, designated V4 Pro 0813, ships at the same price as the preview: $0.435 per million input tokens on a cache miss, $0.87 per million output tokens, and $0.003625 per million on a cache hit. But the pricing page now carries a notice most coverage has missed: DeepSeek plans to raise overall API pricing in the near future, and it describes the increase as significant.

That makes this an unusual moment in token pricing. The list price is known, the increase is announced, and the timing is not. This page covers what V4 Pro costs today, what the increase means, and how buyers can think about the window in between.

DeepSeek V4 Pro official API pricing

Two things stand out. The cache hit discount is over 99 percent, the steepest of any major provider, which means workloads with repetitive context pay a fraction of the list price. And the context window is 1M tokens with no surcharge, at a list price that undercuts most Western flagship models by an order of magnitude.

Both models support JSON output, tool calls, and Anthropic API compatibility.

What changed at GA

On price, nothing. The 0813 build is the first flagship DeepSeek release to exit preview with its preview pricing intact.

On the model itself, V4 Pro is a mixture-of-experts system with 1.6 trillion total parameters and 49 billion active per token. DeepSeek reports that its compressed attention design cuts inference compute to 27 percent of V3.2's at maximum context, with KV cache requirements at 10 percent. Those two engineering numbers are the economic story: they are why a 1M-context flagship can list at $0.435 at all. DeepSeek also reports benchmark results for the Pro Max mode including 80.6 percent on SWE-bench Verified and 93.5 percent on LiveCodeBench.

For how infrastructure efficiency translates into token prices generally, see why token prices differ across providers.

The announced price increase

DeepSeek's official pricing documentation now states that the company plans to raise overall API pricing in the near future, with a significant increase expected. No effective date and no new rates have been published.

This is worth sitting with, because it is a clean illustration of how posted prices work. A posted price is not a market price. It is a number one seller chooses, and it can move by announcement, in either direction, at any time. DeepSeek cut V3.2 prices repeatedly through 2025 and 2026. It is now signaling the opposite move, from a position where its official API is already among the cheapest hosts of its own model.

For buyers, the practical readings are straightforward. Current rates are the floor, not the ceiling, so cost models built on $0.435 input should carry a sensitivity case above it. Cached-input economics may change disproportionately, since the 99 percent cache discount is the most aggressive term on the page. And third-party hosts, who set prices independently, will not necessarily follow the official API upward, which could invert today's ranking where the official API is the cheapest place to run V4.

What a real workload pays today

Take a production workload of 10M input tokens and 1M output tokens per day.

// text
No caching:
10M input  x $0.435  = $4.35
1M output  x $0.87   = $0.87
Total: $5.22/day, $156.60/month, $1,905/year

With a 60% cache hit rate:
4M cache miss  x $0.435    = $1.74
6M cache hit   x $0.003625 = $0.02
1M output      x $0.87     = $0.87
Total: $2.63/day, $78.95/month, $961/year

A 60 percent cache hit rate roughly halves the bill. This is the recurring lesson of DeepSeek pricing analysis: the blended rate, not the list price, is what you actually pay, and cache terms dominate the blend for most production workloads.

V4 Pro across providers

Third-party hosts currently list V4 Pro between $0.59 and $1.74 per million input tokens, all above the official API. The full 16-provider table, and why the developer is the cheapest host this time, is maintained on the DeepSeek V4 API pricing page.

The announced increase makes that spread dynamic. If the official price rises toward the third-party range, the case for the official API weakens and the market re-sorts. Which prices actually move, and when, is exactly the kind of question a posted-price system answers slowly and a live market answers continuously.

How to think about the window

An announced-but-unpriced increase creates a decision window. Locking in current economics is not possible on the official API, which sells at posted prices with no term structure. There is no instrument there to hedge with.

This is the structural gap forward markets exist to fill. When a seller announces "prices will rise significantly" without a date or a number, a buyer with a forward curve can price that risk and lock delivery months ahead. A buyer facing a posted price can only watch the page. The Token Index tracks where V4 Pro and other open-weight models actually trade as this plays out.

Frequently asked questions

What does DeepSeek V4 Pro cost right now?
$0.435 per million input tokens on a cache miss, $0.003625 on a cache hit, and $0.87 per million output tokens, with a 1M token context window at no surcharge.

When do prices increase, and by how much?
Not announced. DeepSeek's pricing page says an increase is planned in the near future and describes it as significant, but publishes no date or new rates. Check the official pricing page before committing to long-term cost models.

Is V4 Pro GA pricing different from preview pricing?
No. The 0813 GA build kept preview pricing unchanged.

Should I use V4 Pro or V4 Flash?
Flash lists at roughly a third of Pro's input price ($0.14 vs $0.435) with the same 1M context window. If your workload doesn't need flagship reasoning quality, Flash is the cost play. Benchmark both on your actual task before deciding on price alone.

Methodology

Official pricing verified against DeepSeek's API documentation on August 13, 2026. Release details from DeepSeek's V4 Pro 0813 general availability announcement. Third-party provider range from Mercatus provider tracking as of the V4 pricing page's last update. Prices change frequently, and DeepSeek has announced a pending increase; treat all figures as a snapshot at the date above.