No description available
DeepSeek V4 Pro APIActive
AI Code With channel availability
The same model is billed by channel discount. Copy a channel code to set your preference.
Related Models
More models from DeepSeek
Compare models from other providers
DeepSeek V4 Pro API: price, model ID, and how to call it
Call DeepSeek V4 Pro through AI Code With. Use model ID deepseek-v4-pro with 1M context. list price about ¥9.00 / ¥27.00 per 1M tokens before channel discounts. About ¥0.6084 for 100,000 input / 10,000 output tokens at the cheapest live channel.
Next: Get API key · See integration guides · Compare DeepSeek models · All model prices
Cost calculator
Estimate at the cheapest live channel: ¥4.68 / ¥14.04 per 1M tokens, cache read ¥0.1560, cache write ¥0.0000. Billed usage follows your prepaid balance.
cURL example
Set model to deepseek-v4-pro. The page slug is only the URL, not the API parameter.
curl https://api.aicodewith.ai/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $AICODEWITH_API_KEY" \
-d '{
"model": "deepseek-v4-pro",
"messages": [
{ "role": "user", "content": "What is the meaning of life?" }
]
}'FAQ
What is the DeepSeek V4 Pro model ID?
Send model "deepseek-v4-pro" in the API request. The page path /deepseek-v4-pro is only the URL, not the model parameter.
How do I call the DeepSeek V4 Pro API?
Use one AI Code With API key against https://api.aicodewith.ai/v1/chat/completions, and set model to deepseek-v4-pro. Supported protocols: OpenAI Chat Completions. See See integration guides.
How is DeepSeek V4 Pro priced?
DeepSeek V4 Pro is list price about ¥9.00 / ¥27.00 per 1M tokens before channel discounts. The public page shows available channel discounts; billed usage follows your prepaid balance and the live channel. About ¥0.6084 for 100,000 input / 10,000 output tokens at the cheapest live channel.
How long is the DeepSeek V4 Pro context window?
DeepSeek V4 Pro has about 1M input context and up to 393.22K output tokens. Treat that as capacity, not a reason to fill every request.
GPT-5.6 Sol