DeepSeek v4.1 Flash

Chat Completions API

deepseek-v4.1-flash

от

DeepseekСоздана 10 сент. 2026 г.1M контекст

Chat Completions API

DeepSeek v4.1 Flash is a sparse mixture-of-experts model from DeepSeek, and the first built on the company's Causal Encoder-Decoder (CED) architecture. It activates 8B parameters on input and 16B on output from a 552B-parameter backbone, an asymmetric split that keeps per-token compute low relative to the model's total size. Image understanding is native to the architecture, with visual and text embeddings trained jointly from the start of pre-training rather than added afterward as in the earlier experimental v4 Flash Vision Exp.

Сравнить Открыть в Плейграунде

Здесь собран публичный контракт интеграции модели: поддерживаемые API-методы, параметры запросов и примеры вызовов.

Chat Completions APIPOST/v1/chat/completions

curl https://api.makura.ai/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  --data-raw '{
  "model": "deepseek-v4.1-flash",
  "messages": [
    {
      "role": "user",
      "content": "Сколько будет 2 + 2?"
    }
  ]
}'

Frequency PenaltyLogit BiasMax Completion TokensParallel Tool CallsPresence PenaltyReasoning EffortResponse FormatStopTemperatureTool ChoiceToolsTop PLogprobsPrediction

DeepSeek v4.1 Flash

API-методы и структура запроса

Пример кода

Параметры