DeepSeek V3
DeepSeek V3 is a text-based AI model developed by DeepSeek, released in December 2024. It is designed to handle complex instruction following and coding tasks with a large 163,840-token context window.
DeepSeek V3 specs and pricing
- Provider: DeepSeek
- Released: December 26, 2024
- Context window: 164K tokens
- Max output: 16K tokens
- Input: text
- Output: text
- Price: $0.26 input / $1.03 output per 1M tokens
- Tool calling: yes
What DeepSeek V3 is good at
- Writing and debugging complex code
- Processing long-form text documents
- Executing multi-step tasks via tool calling
To-dos to try with DeepSeek V3
- Analyze this 50-page technical documentation and summarize the key API requirements.
- Debug this script to identify why the function is returning an unexpected null value.
- Draft a structured project plan based on these meeting transcripts.
FAQ
- What is DeepSeek V3 primarily designed to do? DeepSeek V3 is built to excel at instruction following and coding tasks.
- Does this model support tool calling? Yes, the model is capable of tool calling to assist with various tasks.
- How much does DeepSeek V3 cost? DeepSeek V3 costs $0.26 per million input tokens and $1.03 per million output tokens through OpenRouter. In todo.is a typical agent step costs about 0.86 credits ($1 = 100 credits).
- What is the context window of DeepSeek V3? DeepSeek V3 can read up to 164K tokens at once and write up to 16K tokens in one reply.
- Can I use DeepSeek V3 without coding? Yes. In todo.is you pick DeepSeek V3 as your agent's model and write to-dos in plain words, on the web, the Android app, WhatsApp or Telegram. No API key needed: credits are included, or you can bring your own key.
JavaScript is required to use the todo.is app.