Tencent: Hunyuan A13B Instruct

Tencent's Hunyuan-A13B Instruct is a powerful 13B active parameter Mixture-of-Experts (MoE) language model, boasting an impressive total parameter count of 80B. Designed for advanced applications, it supports sophisticated reasoning via Chain-of-Thought, making it highly effective for complex problem-solving. This model demonstrates competitive benchmark performance across a wide range of domains, including mathematics, science, coding, and multi-turn reasoning tasks. It achieves high inference efficiency through Grouped Query Attention (GQA) and supports various quantization methods like FP8 and GPTQ. With a generous context window of 131K tokens and a max output of 4K tokens, it's ideal for extensive conversational and generative tasks. Pricing is set at $0.14 per 1M input tokens and $0.57 per 1M output tokens, available at the STARTER access tier. It supports code generation and streaming capabilities.