Skip to main content
Back to models

Model comparison

Compare model identity and capabilities.

Selected

DeepSeek V4 Flash 0731

deepseek/deepseek-v4-flash-0731

DeepSeek / DeepSeek

View model

DeepSeek V4 Flash 0731 is a sparse mixture-of-experts model from DeepSeek, with 13B active parameters out of 284B total. This re-post-trained revision is suited for coding, reasoning, and agent workflows.

Released
2026-07-31
Context window
1M
Max output
66K
Weights
Open weight
Modalities
TextInputTextOutput
Capabilities
ReasoningTool callStructuredTemperatureOpen weight
Providers
AmbientKilo GatewayNanoGPTOpenRouterTensorXVercel AI Gateway

Select at least two models to compare.

Compare models
1 models selected
Compare