The Daily Commit · Section Edition Front Page PHP AI Dev EN DE FR ES

TheModelDesk

August 12, 2026
models, agents & local inference

Releases

DeepSeek Releases V4 Pro 0813 Mixture-of-Experts Model

DeepSeek has released V4 Pro 0813, a large-scale mixture-of-experts language model available via OpenRouter. The model supports 1M token context length and is priced at $0.435 per million input tokens and $0.87 per million output tokens. Released August 12, 2026, it offers general availability with OpenAI-compatible API access.

DeepSeek announced the general availability release of V4 Pro 0813, a large-scale mixture-of-experts model designed for production workloads.

The model features a 1-million-token context window and is available through OpenRouter's API platform. Pricing is set at $0.435 per million input tokens and $0.87 per million output tokens, reflecting competitive positioning in the market for efficient inference.

V4 Pro 0813 launched on August 12, 2026. The model supports OpenAI-compatible API calls, enabling developers to integrate it using standard SDKs with only a base URL change. OpenRouter provides monitoring infrastructure, performance metrics (throughput, latency, time-to-first-token), and benchmark comparisons against other hosted models.

The release includes standard capabilities such as tool calling and structured output support. The mixture-of-experts architecture allows the model to optimize computation by activating only relevant expert networks for each task, a design pattern increasingly adopted across large language models for efficiency gains.

Read the original source ↗

Rate this article: 0

Readers’ Forum

No contributions yet — open the debate.

← The Model Desk — Page C1

Models, agents & local inference · The Daily Commit · Screen edition · Imprint · Privacy Policy