Qwen3-Omni-Flash
Qwen3-Omni-Flash is an omni-modal model based on a Thinker–Talker mixture-of-experts architecture. This ID identifies the September 15, 2025 snapshot; Alibaba listed it in its release notes on…
MODEL DIRECTORY
New releases, in one place. Explore models by their publication date.
Qwen3-Omni-Flash is an omni-modal model based on a Thinker–Talker mixture-of-experts architecture. This ID identifies the September 15, 2025 snapshot; Alibaba listed it in its release notes on…
Qwen3-Omni-Flash-Realtime is the real-time version of Qwen3-Omni-Flash, based on a Thinker–Talker mixture-of-experts architecture. This model ID identifies the September 15, 2025 snapshot. Alibaba’s…
Tencent’s Hunyuan 3D 3.0 is a generative model for creating 3D assets. Tencent states that its Hunyuan 3D Professional service launched on September 16, 2025, using the 3.0 model to produce…
Released: September 15, 2025. GPT-5-Codex is a version of GPT-5 optimized for agentic coding in Codex and similar environments. (openai.com) The model was designed for interactive code edits, longer…
Qwen-VL-OCR is an optical character recognition model trained on Qwen-VL. It brings image-text recognition, parsing, and processing tasks together in one model. It is intended for extracting and…
Wan2.2-KF2V-Flash is a Wan 2.2 first-and-last-frame video generation model. It takes two images and a text prompt and generates a video transitioning between the images. Alibaba describes improvements…
MiniMax Music 1.5 is a text-guided music generation model announced on September 11, 2025. It generates songs up to four minutes long from natural-language descriptions and can use lyrics arranged for…
Qwen-Plus is a text-generation and reasoning model. This dated snapshot was released on September 11, 2025. Alibaba reports improved instruction following and more concise summaries in thinking mode,…
ERNIE-X1.1 is Baidu’s deep-thinking reasoning model, released on September 9, 2025. Baidu describes it as an upgrade to ERNIE X1 using an iterative hybrid reinforcement-learning framework and…
Qwen3-ASR-Flash is a multilingual automatic speech recognition model based on a large language model. This model ID identifies the September 8, 2025 snapshot. It takes audio as input and returns text.…
Kimi-K2-Instruct-0905 is an instruction-tuned Kimi K2 language model aimed at coding and agentic tasks. It uses a mixture-of-experts architecture with 1 trillion total parameters and 32 billion…
Qwen3-Max-Preview is a preview of the Max model in the Qwen3 series, positioned for text generation and reasoning. Alibaba’s release record dates it to September 5, 2025. (help.aliyun.com) Alibaba…
CosyVoice-v3-plus is a Tongyi Lab speech-synthesis and voice-cloning model intended for professional applications. It converts text to audio and supports real-time streaming text-to-speech. Alibaba…
Qianfan-EngCard-VL is Baidu's domain-specialized image understanding model for educational scenarios. Baidu Qianfan lists it as an image understanding model in the Inference Service API V2.…
A domain-specialized image understanding model from Baidu Qianfan, designed to recognize and extract information from a single image; the official community lists it as a model for the education…
Wan2.2-S2V-Detect is an auxiliary image detection model for the Wanxiang digital human video model wan2.2-s2v. It checks whether a person's portrait meets that model's input specifications before…
Qwen-Deep-Research (Qwen Deep Research) is an advanced agent system for complex research tasks. The official release announcement was dated 2025-08-22. qwen-deep-research is a continuously updated…
Qwen-MT-Image is an image translation model service provided by Alibaba Cloud Model Studio. It takes an image as input and outputs a translated image, targeting image localization scenarios. It was…
An introduction in this language is coming soon.
DeepSeek-V3.1-Base is the base pretrained language model for DeepSeek-V3.1. It was further trained for long-context extension on the original DeepSeek-V3 base model checkpoint; the post-trained…