DeepSeek V4 Flash is a hybrid-thinking model with 1M context and 384K max output. It supports both non-thinking and thinking modes, with thinking enabled by default.
Specifications
Context
1000K
Maximum Output
384K
Inputtext
Outputtext
Performance (7-day Average)
Collecting…
Collecting…
Collecting…
Pricing
Input¥1.00/MTokens
Cached Input¥0.20/MTokens
Output¥2.00/MTokens
Availability Trend (24h)
Performance Metrics (24h)
Similar Models
¥1.00/¥2.00/M
ctx1.0Mmax384Kavail—tps—
InOutCap
Compatibility alias for deepseek-v4-flash in non-thinking mode. DeepSeek will deprecate this alias in the future.
¥1.00/¥2.00/M
ctx1.0Mmax384Kavail—tps—
Compatibility alias for deepseek-v4-flash in thinking mode. DeepSeek will deprecate this alias in the future.
¥12.00/¥24.00/M
ctx1.0Mmax384Kavail—tps—
InOutCap
DeepSeek V4 Pro is the higher-capability hybrid-thinking model in the V4 family. It supports both non-thinking and thinking modes, with 1M context and 384K max output.
$2.20/$8.80/M
ctx1.0Mmax33Kavail—tps—
InOutCap
GPT-4.1 is an enhanced version of GPT-4 with improved instruction following and multimodal capabilities for text and image understanding.