HunYuan T1

An industry-leading deep reasoning large model, optimized for human preferences.

ChineseSelectionProductivityDeep LearningReasoning Model

HunYuan T1 is a deep reasoning large model based on reinforcement learning, launched by Tencent. Through extensive post-training and alignment with human preferences, it significantly improves reasoning ability and efficiency. The product is based on a large-scale Hybrid-Transformer-Mamba MoE architecture, enabling the model to perform better when handling long texts. Suitable for various users who need complex reasoning and logical solutions, assisting scientific research and technological development.

Best AI Websites & Tools

HunYuan T1

HunYuan T1 Alternatives

HunYuan T1 — The industry's first ultra-large-scale hybrid Mamba reasoning model, with strong reasoning capabilities.

HunYuan T1 — An industry-leading deep reasoning large model, optimized for human preferences.

R1-Omni — R1-Omni is a full-modality emotion recognition model incorporating reinforcement learning, focusing on improving the interpretability of multimodal emotion recognition.

Steiner-32b-preview — Steiner is a reasoning model trained on synthetic data, designed to explore multiple reasoning paths and verify them autonomously.

VLM-R1 — VLM-R1 is a stable and versatile reinforcement learning-enhanced visual-language model focused on visual understanding tasks.

DeepSeek-R1-Distill-Qwen-7B — DeepSeek-R1-Distill-Qwen-7B is an open-source reasoning model focusing on mathematics, coding, and reasoning tasks.

mwp_ReFT — A deep reinforcement learning-based model fine-tuning framework

DeepMind — A leading artificial intelligence research company under Google

DiffusionRL — Large-scale Reinforcement Learning for Diffusion Models

RF-DETR — RF-DETR is a real-time object detection model developed by Roboflow.

InfiniteYou — Achieve flexible and high-fidelity image generation while preserving identity characteristics.

Reka Flash 3 — A 21B general-purpose reasoning model suitable for low-latency applications.

Pruna — Pruna is a model optimization framework that helps developers deliver models quickly and efficiently.

EXAONE Deep — An open-source reasoning AI model launched by LG AI, boasting excellent reasoning capabilities.

Long Context Tuning (LCT) — A technology that enhances scene-level video generation capabilities.

Thera — An aliasing-free arbitrary-scale super-resolution method.

Light-R1-14B-DS — An open-source 14B-parameter mathematical model, trained using reinforcement learning, with excellent performance.

Light-R1 — Light-R1 is an open-source project focusing on long-chain reasoning (Long COT), providing a training method from scratch through curriculum-style SFT, DPO, and RL.

IMM — Inductive Moment Matching is a novel generative model for high-quality image generation.

MIDI — Generates high-fidelity 3D scenes from a single image using a multi-instance diffusion model.

VideoPainter — VideoPainter is a tool that supports video repair and editing of any length, using a text-guided plug-in framework.

Bytedance Flux — Flux is a fast communication overlap library for tensor/expert parallelism on GPUs.

NotaGen — NotaGen is a model for symbolic music generation, employing a large language model training paradigm and focusing on generating high-quality classical music scores.

HunyuanVideo-I2V — HunyuanVideo-I2V is an image-to-video generation framework based on HunyuanVideo, launched by Tencent.

QwQ-32B — QwQ-32B is a powerful reasoning model designed for complex problem-solving and text generation, delivering exceptional performance.

CogView4-6B — CogView4-6B is a powerful text-to-image generation model focusing on high-quality image generation.

UniTok — UniTok is a unified visual tokenizer for visual generation and understanding.

PhotoDoodle — PhotoDoodle is a code implementation of artistic image editing based on learning from a small number of sample pairs.

Profiling Data in DeepSeek Infra — Analyzes the computation and communication overlap strategies in V3/R1, providing performance analysis data for deep learning frameworks.

EPLB — An open-source algorithm for expert parallelism load balancing, designed to optimize expert allocation and load balancing in multi-GPU environments.