Back to Dashboard
💬 Text ModelOpen Source
DeepSeek V4 Pro
by DeepSeek
Benchmark Evaluations
Codingswe-bench-pro
52.1%
Reasoninggpqa-diamond
88%
Pricing & Speed Details
Input Cost
$1.74
per 1M tokensOutput Cost
$3.48
per 1M tokensBlended Cost
$2.17
based on 3:1 ratioSpeed
39.7 tps
Context Window
1M tokens
Want to test this model?
Compare DeepSeek V4 Pro dynamically against other top AI models on our comparison dashboard to optimize your choice.
Go to DashboardQuick Info
- ProviderDeepSeek
- CategoryText / Chat Generation
- LicenseOpen Source
Frequently Asked Questions
DeepSeek V4 Pro has a blended cost of $2.17 per 1 million tokens (calculated using a standard 3:1 input-to-output token ratio). Its specific pricing is $1.74 per 1M input tokens and $3.48 per 1M output tokens.
DeepSeek V4 Pro supports a context window of 1M tokens. This allows it to process large documents, codebase segments, or long chat histories in a single query session.
DeepSeek V4 Pro scores 52.1% on coding evaluations (SWE-bench-pro) and 88% on reasoning tasks (GPQA-diamond).
DeepSeek V4 Pro is an open source model developed by DeepSeek.