DeepSeek V4 Pro Quietly Released: Intelligence Index 53, Above-Average Pricing

ns123abc · x · 2026-08-13

DeepSeek made no major announcement for V4 Pro on social media, disappointing some users as it scores only 1 point higher than the Flash version. According to Artificial Analysis, DeepSeek V4 Pro 0813 (Reasoning, Max Effort) achieves an Intelligence Index of 53, well above the median of 27.

Performance-wise, the model delivers 83 output tokens per second (faster than average) and supports a 1M token context window. Pricing is set at $0.435 per 1M input tokens and $0.87 per 1M output tokens, making it somewhat expensive compared to similar open-weight models, with a tendency to be verbose.

Related event: DeepSeek V4 Pro Launch Sparks Debate: Modest Gains, Strong Security(12 posts)→

Original post →

More from Models

Models channel →