DeepSeek V4 Pro 0813 Released on OpenRouter
Original: DeepSeek V4 Pro 0813
Why This Matters
The GA release of DeepSeek V4 Pro marks a significant milestone in large-scale open MoE model availability for enterprise API use.
DeepSeek has released DeepSeek V4 Pro 0813, a large-scale mixture-of-experts (MoE) model, on OpenRouter as of August 12, 2026. Pricing is set at $0.435 per 1M input tokens and $0.87 per 1M output tokens, with a 1M token context window.
DeepSeek has officially launched DeepSeek V4 Pro 0813, described as the GA (General Availability) release of DeepSeek V4 Pro. The model is a large-scale mixture-of-experts (MoE) architecture and is available via OpenRouter's API platform. It was released on August 12, 2026, and is currently hosted by a single provider on OpenRouter. Pricing is listed at $0.435 per 1 million input tokens and $0.87 per 1 million output tokens, though OpenRouter notes that actual costs paid by customers are often lower due to caching and available discounts. The model supports a 1 million token context window and is compatible with OpenAI-style API calls, requiring only a model slug change for integration. OpenRouter monitors provider uptime continuously and routes retries automatically in case of errors.