DeepSeek-V3.1 (deepseek-chat) is a powerful generalist model with 671B parameters, offering exceptional performance at an economical price. It achieves strong results on mathematical reasoning, coding, and general language tasks. The model supports 128K context length with a default output of 4K tokens (max 8K) and features advanced capabilities like function calling and JSON output.
Input: Text
Output: Text
Providers
deepinfra
Credits
Context128k
Max Output8k
Input$0.270/1M
Output$1.00/1M
Cache Read$0.216/1M
Cache Write—
deepseek
Credits
Context128k
Max Output8k
Input$0.560/1M
Output$1.68/1M
Cache Read$0.070/1M
Cache Write—
openrouter
Credits
Context128k
Max Output8k
Input (Max)$3.16/1M
Output (Max)$4.75/1M
Cache Read—
Cache Write—
Quick Start
Use DeepSeek V3 through Helicone's AI Gateway with automatic logging and monitoring.