Pricing updated 2026-08-19: AIWave V4 Flash is $0.638 input, $1.914 output, and $0.0203 cache hit; V4 Pro is $1.914 input, $5.742 output, and $0.0638 cache hit per 1M tokens.
title: "DeepSeek API Pricing 2026: V4 Flash, V4 Pro and Cost Breakdown"
published: true
tags: deepseek, pricing, api-cost, deepseek-v4
canonical_url: https://aiwave.live/blog/deepseek-api-pricing-2026
description: "DeepSeek API pricing updated 2026-08-19. AIWave V4 Flash is $0.638/$1.914 and V4 Pro is $1.914/$5.742 per 1M tokens."
DeepSeek offers two flagship models in 2026: V4 Flash (speed and cost) and V4 Pro (quality). Both are available through AIWave at a small markup over direct DeepSeek pricing, with the benefit of a single API key and PayPal payments.
| DeepSeek Direct | AIWave | Markup | |
|---|---|---|---|
| V4 Flash Input | $0.44 peak | $0.638 all day | Gateway rate |
| V4 Flash Output | $1.32 peak | $1.914 all day | Gateway rate |
| V4 Pro Input | $1.32 peak | $1.914 all day | Gateway rate |
| V4 Pro Output | $3.96 peak | $5.742 all day | Gateway rate |
DeepSeek direct prices from api-docs.deepseek.com. AIWave prices from aiwave.live/pricing.
AIWave is actually slightly cheaper than going direct for V4 Pro. This is possible because AIWave negotiates volume pricing and the small overhead from their aggregation layer is absorbed.
| Model | Context | Max Output | Cost per full context fill |
|---|---|---|---|
| V4 Flash | 1M tokens | 384K tokens | $0.31 |
| V4 Pro | 1M tokens | 384K tokens | $0.93 |
| GPT-4o (comparison) | 128K tokens | 16K tokens | $0.42 |
DeepSeek's 1M context window is 5-8x larger than most competitors. For codebase analysis and long documents, this means fewer API calls to chunk and reassemble content.
200 calls/day, 4K input + 2K output, 22 days:
| Model | Monthly |
|---|---|
| V4 Flash | $5 |
| V4 Pro | $18 |
| GPT-4o | $132 |
| Claude Sonnet 4 | $185 |
| Model | Monthly |
|---|---|
| V4 Flash | $1.65 |
| V4 Pro | $8 |
| GPT-4o | $59 |
DeepSeek offers cached input discounts (prefix matching). When your request shares a prefix with a recent request, input tokens are cheaper:
| Normal Input | Cached Input | |
|---|---|---|
| V4 Flash (AIWave) | $0.638/1M | $0.0203/1M |
| V4 Pro (AIWave) | $1.914/1M | $0.0638/1M |
Cached input works automatically — no code changes needed. Repeated prompts or system instructions trigger the discount.
AIWave provides DeepSeek V4 Flash as deepseek-chat and V4 Pro as deepseek-reasoner. Same OpenAI-compatible format, same response structure. The $0.20 starter credit covers roughly 700 V4 Flash coding queries — enough for two weeks of daily use.