DeepSeek-V4.1-Flash-1017 is a minor update to DeepSeek-V4.1-Flash from DeepSeek. It uses the same multimodal Mixture-of-Experts (MoE) architecture, accepts text and image input, and has a context window of up to 1M tokens. DeepSeek reports improved results on reasoning, coding, agentic, and visual benchmarks compared to DeepSeek-V4.1-Flash, along with shorter outputs and faster task completion.
The deepseek-v4.1-flash-1017-maas endpoint is available until at
least December 31, 2026. Availability might be extended based on usage and
demand. For more information, see Minimum availability for Preview
models.
Managed API (MaaS) specifications
| Model ID | deepseek-v4.1-flash-1017-maas |
|
|---|---|---|
| Modalities |
|
|
| Capabilities |
|
|
| Consumption options |
|
|
| Supported regions |
|
|
|
||
| Quotas |
|
|
| Versions |
|
|
Deploy as a self-deployed model
To self-deploy the model, go to the DeepSeek-V4.1-Flash-1017 model card in the Model Garden console and click Deploy model. For more information about deploying open models, see Deploy open models from Model Garden.
What's next
- Before you use the model, grant user access to open models.
- Learn how to call MaaS APIs for open models.
- Learn how to use function calling for open models.
- Learn how to use structured output for open models.