DeepSeek Models

On October 17, 2026, the company released **DeepSeek-V4.1-Flash-1017**, a targeted update to its V4.1-Flash model (originally launched September 10, 2026). Marketed as delivering substantial improvements in reasoning, coding, agentic capabilities, and visual understanding—while reducing task execution time and output length—this checkpoint is positioned as DeepSeek’s most powerful efficient model to date for real-world agent and coding workloads.

Explore the DeepSeek V4.1 Pro API.

Discover DeepSeek Flash, a 552B multimodal MoE model with 1M-token context, adjustable reasoning, benchmark results, pricing, and CometAPI access.

Модель deepseek-v4-flash-vision-exp принимает изображения вместе с текстом, поэтому вы можете просить модель описывать изображения, считывать текст с снимков экрана, анализировать диаграммы и многое другое. Поддерживаемые форматы изображений: JPEG, PNG, GIF и WebP. Формат определяется по фактическому содержимому файла, а не по имени файла или заявленному типу MIME.

DeepSeek V4 Pro is a large-scale Mixture-of-Experts model from DeepSeek with 1.6T total parameters and 49B activated parameters, supporting a 1M-token context window. It is designed for advanced reasoning, coding, and long-horizon agent workflows, with strong performance across knowledge, math, and software engineering benchmarks.

deepseek-r2 coming soon

DeepSeek V4 Flash is an efficiency-optimized Mixture-of-Experts model from DeepSeek with 284B total parameters and 13B activated parameters, supporting a 1M-token context window. It is designed for fast inference and high-throughput workloads, while maintaining strong reasoning and coding performance.

The most popular and cost-effective DeepSeek-V3 model. 671B full-blood version. This model supports a maximum context length of 64,000 tokens.