DeepSeek V4 Pro 0423 API Prices Rise 51% — What Developers Should Know

The Numbers

DeepSeek has pushed through a uniform +51% price increase on its V4 Pro 0423 API. Input tokens move from $0.522 to $0.79 per 1M tokens; output tokens go from $1.04 to $1.58 per 1M tokens. Both sides of the pricing ledger take the same proportional hit, so the input/output ratio stays roughly consistent.

Does This Matter for Your Workload?

At the old rates, a workload burning 10M input tokens and 5M output tokens per month cost around $10.42. At the new rates, that same workload runs approximately $13.10 — a real but not catastrophic jump for moderate usage. Heavy users running hundreds of millions of tokens monthly will feel this more sharply; the math scales linearly.

For context, DeepSeek V4 Pro was already positioned as a value option relative to flagship models from OpenAI and Anthropic. Even at $0.79 input / $1.58 output, it remains competitive in that tier — but the gap has narrowed, and the increase gives legitimate reason to re-evaluate alternatives.

What to Do Next

• Audit your token split. Output-heavy pipelines (long-form generation, chain-of-thought) take a harder hit at $1.58/1M than input-heavy ones.
• Re-benchmark your use case. If you've been coasting on old pricing assumptions, update your cost models now.
• Compare before committing. Check DeepSeek V4 Pro 0423 — live specs & price history for up-to-date figures and side-by-side model comparisons on our site.

A 51% increase is significant enough to warrant a genuine review — not panic, but not inaction either.

Ezra, Scout AI Team