Model · DeepSeek

DeepSeek V4 Flash

deepseek/deepseek-v4-flash

Try in Agents

DeepSeek V4 fast variant with hybrid sparse attention and stronger agentic tool use

cachinglegacy

Specs

Context

1.0M

Max output

393K

Provider

DeepSeek

Input

$0.076/ 1M tokens

Output

$0.153/ 1M tokens

Cache read

$0.014/ 1M tokens

Try it

Chat with DeepSeek V4 FlashTerramind API

⌘+Enter to run

Code

import { streamText } from "ai";
import { createOpenAICompatible } from "@ai-sdk/openai-compatible";

const terramind = createOpenAICompatible({
  name: "terramind",
  baseURL: "https://terramind.com/api/v1",
  apiKey: process.env.TERRAMIND_API_KEY!,
});

const result = streamText({
  model: terramind("deepseek-v4-flash"),
  prompt: "Why is the sky blue?",
});

for await (const chunk of result.textStream) {
  process.stdout.write(chunk);
}

Want to run this in your app? Grab an API key · Full chat reference

More by DeepSeek

Model	Context	In	Out	Capabilities
deepseek/deepseek-v4-flash-vision	1M	$0.220	$0.660	visioncaching
deepseek/deepseek-v4-pro	1.0M	$0.660	$1.98	reasoningcaching
deepseek/deepseek-v4-1-flash	1.0M	$0.300	$1.20	visioncaching