DeepSeek-V4.1-Flash

DeepSeek-V4.1-Flash is a multimodal Mixture-of-Experts (MoE) model from DeepSeek built for agentic and coding tasks. It accepts text and image input and has a context window of up to 1M tokens. It activates 8B of its 552B backbone parameters per token during prefill, which keeps input-heavy agentic workloads cost-efficient.

The deepseek-v4.1-flash-maas endpoint is available until at least November 28, 2026. Availability might be extended based on usage and demand. For more information, see Minimum availability for Preview models.

Managed API (MaaS) specifications

Pricing

Model ID deepseek-v4.1-flash-maas
Modalities
Text Input and output
Image Input only
Audio Not supported
Video Not supported
Capabilities
Consumption options
Supported regions

Model availability

  • Global: global

ML processing

  • Global: global
Quotas
  • global: 1,048,576 maximum output, 1,048,576 context length
Versions
  • deepseek-v4.1-flash-maas
    • Launch stage: Preview
    • Release date: October 14, 2026

Deploy as a self-deployed model

To self-deploy the model, navigate to the DeepSeek-V4.1-Flash model card in the Model Garden console and click Deploy model. For more information about deploying and using partner models, see Deploy a partner model and make prediction requests.