DeepSeek releases V3.1 model, supporting proxy and hybrid reasoning

V3.1-Think mode inference speed is improved compared to DeepSeek-R1-0528, and it enhances tool usage and multi-step proxy task capabilities. The API now supports 128KB context, compatible with the Anthropic API format and strict function calls. The V3.1 base model extends long context with an 840KB token. Open source weights are now available, and pricing will be adjusted on September 5th.