DeepSeek: DeepSeek V4 Flash

Text input Text output
Author's Description

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...

Key Specifications
Cost
$$$
Context
1M
Parameters
284B (Rumoured)
Released
Apr 23, 2026
Speed
Ability
Reliability
Supported Parameters

This model supports the following parameters:

Top P Presence Penalty Top Logprobs Include Reasoning Temperature Response Format Logprobs Reasoning Tools Tool Choice Max Tokens Frequency Penalty Stop
Features

This model supports the following features:

Tools Reasoning Response Format
Performance Summary

DeepSeek V4 Flash, an efficiency-optimized Mixture-of-Experts model with 284B total and 13B activated parameters, demonstrates a balanced performance profile. Its speed performance is moderate, ranking in the 36th percentile across benchmarks. However, it offers competitive pricing, typically providing cost-effective solutions and ranking in the 61st percentile. A standout feature is its exceptional reliability, achieving a 100% success rate across all benchmarks, indicating consistent and stable operation. The model exhibits strong accuracy in several key areas. It achieved perfect scores in Reasoning and Ethics, with the Reasoning benchmark specifically noting it as the most accurate model at its price point and speed. Instruction Following, Coding, Email Classification, and Mathematics also show high accuracy, ranging from 84% to 99%. Its General Knowledge is robust at 99.5%. A relative weakness is observed in Hallucinations, where its 92% accuracy, while good, places it in the 42nd percentile, suggesting some room for improvement in acknowledging uncertainty. Overall, DeepSeek V4 Flash is a highly reliable and cost-effective model with strong performance across most cognitive tasks, particularly excelling in complex reasoning and ethical considerations.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $0.14
Completion $0.28
Input Cache Read $0.0028

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
DeepSeek
DeepSeek | deepseek/deepseek-v4-flash-20260423 1M $0.14 / 1M tokens $0.28 / 1M tokens
DeepInfra
DeepInfra | deepseek/deepseek-v4-flash-20260423 1M $0.14 / 1M tokens $0.28 / 1M tokens
Novita
Novita | deepseek/deepseek-v4-flash-20260423 1M $0.14 / 1M tokens $0.28 / 1M tokens
SiliconFlow
SiliconFlow | deepseek/deepseek-v4-flash-20260423 1M $0.14 / 1M tokens $0.28 / 1M tokens
AkashML
AkashML | deepseek/deepseek-v4-flash-20260423 262K $0.14 / 1M tokens $0.28 / 1M tokens
Parasail
Parasail | deepseek/deepseek-v4-flash-20260423 1M $0.14 / 1M tokens $0.28 / 1M tokens
AtlasCloud
AtlasCloud | deepseek/deepseek-v4-flash-20260423 1M $0.14 / 1M tokens $0.28 / 1M tokens
Venice
Venice | deepseek/deepseek-v4-flash-20260423 1M $0.17 / 1M tokens $0.35 / 1M tokens
Alibaba
Alibaba | deepseek/deepseek-v4-flash-20260423 1M $0.2 / 1M tokens $0.4 / 1M tokens
Novita
Novita | deepseek/deepseek-v4-flash-20260423 1M $0.14 / 1M tokens $0.28 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Strategy Free Executions Accuracy Cost Duration
Other Models by deepseek