Vendor · DeepSeek · 2 models

DeepSeek API on Crazyrouter

DeepSeek's own API is cheap but its payment options and uptime under load are the usual complaints. Crazyrouter routes V4 Pro and V4 Flash across several upstream channels with failover, prices them at or below DeepSeek list, and lets you switch to Claude or GPT for the hard turns without a second account.

Cheapest DeepSeek model here: DeepSeek V4 Flash at $0.45 / $1.34 per 1M tokens.

Get an API key →Price comparisonOfficial docs: deepseek.com
Input price per 1M tokens: list vs Crazyrouter
ListCrazyrouterDeepSeek V4 Pro$1.32 → $1.34DeepSeek V4 Flash$0.300 → $0.450
DeepSeek on Crazyrouter

Chat and reasoning models

DeepSeek
Input / 1M$1.32$1.34
Output / 1M$3.96$4.00
Context 1M tokensCached $0.050Also on Azure
DeepSeek V4 Prodeepseek-v4-pro
Crazyrouter price: $1.34 / $4.00 per 1M tokens · list $1.32 / $3.96
OpenAI compatibleReasoningCoding
Details, samples and code →Compare prices
DeepSeek
Input / 1M$0.300$0.450
Output / 1M$1.20$1.34
Context 1M tokensCached $0.020Also on Azure
DeepSeek V4 Flashdeepseek-v4-flash
Crazyrouter price: $0.45 / $1.34 per 1M tokens · list $0.30 / $1.20
OpenAI compatibleFastLow cost
Details, samples and code →Compare prices

Why route DeepSeek through Crazyrouter

  • V4 Pro at $0.67 / $2.00 per 1M tokens — below DeepSeek's $1.32 / $3.96 list.
  • Failover across channels when DeepSeek's own endpoint is saturated.
  • Same OpenAI-compatible request shape, so `model: "deepseek-v4-pro"` is the only change.
  • Mix with Claude / GPT / Gemini in one bill for router-style setups.

FAQ

Is DeepSeek cheaper here than on DeepSeek's site?
For V4 Pro yes ($0.67 / $2.00 vs $1.32 / $3.96); V4 Flash is at list. Prices are live on each card.
Does reasoning output work?
Yes, `reasoning_content` is returned as DeepSeek documents it.

Further reading