API Key — obtain from https://platform.deepseek.com/

DeepSeek

API model; setup uses a conservative output budget.

deepseek-flash

DeepSeek V3.2 non-thinking mode — general tasks, tool use, 8K max output

DeepSeek V3.2 thinking mode — Chain-of-Thought reasoning, 64K max output

Mit der Modell-API von DeepSeek verbinden.

Conectar con la API de modelos de DeepSeek.

Se connecter à l’API de modèles de DeepSeek.

DeepSeek のモデル API に接続します。

API Type	`openai-completions`
Base URL	`https://api.deepseek.com`
Auth Mode	api_key
Auth Keys	`DEEPSEEK_API_KEY` — API Key — obtain from https://platform.deepseek.com/
Regions	cn, us
Default Model	`deepseek-flash`
Status	active

Model	Context	Max Output	Pricing
deepseek-flash `deepseek-flash` API model; setup uses a conservative output budget.	1.0M	384K	See official pricing documentation.
DeepSeek V4 Pro `deepseek-v4-pro` Reasoning and tool-use model.	1.0M	384K	See official pricing documentation.
DeepSeek V3.2 (Chat) `deepseek-chat` DeepSeek V3.2 non-thinking mode — general tasks, tool use, 8K max output	131K	8K	Input $0.28/M, output $0.42/M (cache hit: input $0.07/M).
DeepSeek V3.2 (Reasoner) `deepseek-reasoner` DeepSeek V3.2 thinking mode — Chain-of-Thought reasoning, 64K max output	131K	64K	Input $0.28/M, output $0.42/M (cache hit: input $0.07/M).