DeepSeek 4 Flash API

A 284B-parameter Mixture-of-Experts model engineered for fast, affordable inference without sacrificing reasoning depth. Thirteen billion parameters active per forward pass. One million tokens of context.
Output

How to use DeepSeek 4 Flash API

Install any OpenAI-compatible SDK, point it at api.aimlapi.com/v1, and set the model to .

OpenAI-compatible — swap the base URL and it works with your existing SDK.

DeepSeek 4 Flash API Pricing

TypePrice
Input
Output

DeepSeek 4 Flash Benchmarks

BenchmarkScoreWhat it measuresSourceRetrieved
Terminal-Bench
82.7
Autonomous shell/terminal task completionSourceSeptember 22, 2026
GPQA Diamond
94.3
Google-proof graduate science questions (hardest subset)SourceSeptember 22, 2026
LiveCodeBench
93.5
Contamination-free competitive programming problemsSourceSeptember 22, 2026

Frequently asked questions

Yes, DeepSeek 4 Flash can stream responses as they are generated.

Yes, DeepSeek 4 Flash supports both function calling and structured outputs.

DeepSeek 4 Flash was built by DeepSeek.

Start building with DeepSeek 4 Flash

Get API Key
1000+ models, one API.