SuperAPI AI Directory
Partner中文AI newsFavorites

All AI tools

35,738 results · Page 354 of 993

Qwen Image 3.0

AI images

Qwen Image 3.0 is an AI image model focused on text rendering, layout, and multilingual design. It can transform structured prompts into image drafts centered on text and layout, such as posters, infographics, product visuals, multilingual design mockups, and storyboards. Users need to describe the purpose, subject, canvas and layout, exact text, visual style, and constraints, select a 1:1, 16:9, or 9:16 canvas, then generate an initial draft, and review the text, numbers, translations, and details one by one. The model supports longer and structured instructions, denser layouts, rendering of smaller text, multilingual design, realistic details, UI simulation, and broader visual knowledge. Text rendering is its key capability, but all results should be treated as drafts, and spelling, numbers, formulas, and factual claims must be verified before publication. Commercial use depends on the terms of the model service used, the source of materials, and applicable law.

Read more

Qwen Image X

AI images

Qwen Image X is a free online Qwen Image AI image generator focused on the Qwen Image series of models (including Qwen Image 3). Users can generate images from text or perform image-to-image editing in the browser, with support for dense layouts, small text of about 10px, and native text rendering in 12 languages, making it suitable for scenarios such as posters, infographics, exam papers, storyboards, and packaging mockups. It offers a daily free quota, no credit card required, and paid plans include commercial licensing.

Read more

qwen-dianjin

AI finance

Qwen DianJin is a collection of open-source projects from the Alibaba Cloud financial services industry team for exploration, research, and deployment of large models in the financial domain. The repository is organized around the financial data flywheel methodology of "evaluation → data synthesis → post-training," and has released achievements including financial evaluation benchmarks, training datasets, reasoning models, process reward models, multimodal document understanding models, and a financial agent skill library. Representative outputs include the CFLUE Chinese financial language understanding evaluation dataset, DianJin-R1-7B/13B and DianJin-R1-Data, the Fin-PRM financial-domain process reward model, DianJin-OCR-R1, and others; it also covers directions such as M³FinMeeting, FinMCP-Bench, DianJin-RED, and DianJin-SKILLS (Baiji Tu). The research is productized on the Alibaba Cloud Qwen DianJin platform. The code and documentation are open under the MIT license, but some components and datasets have separate license terms.

Read more

Qwen3 TTS

AI audio

Qwen3 TTS is a multilingual AI text-to-speech tool based on the Qwen3 TTS model, focused on ultra-fast processing speed. The official claim is that first-packet processing takes only 97ms, making it suitable for real-time speech synthesis scenarios. It offers a free browser demo that can be tried without registration, supports 17 voices and 10 languages, and has Chinese dialect synthesis capabilities. Users can choose built-in voices, clone their own voice, or design new voices. The official demo limits each generation to 120 characters. The model is open on Hugging Face as Qwen/Qwen3-TTS-Demo, with technical documentation and examples provided. The site also presents multiple subscription plans covering different credits, character limits, and commercial usage rights, among other things; specific prices are not provided.

Read more

Qwen3-ASR

AI audio

Qwen3-ASR is an open-source automatic speech recognition (ASR) model series developed by Alibaba Cloud's Tongyi Qianwen team, supporting stable recognition of multilingual speech, music, and songs, language identification, and timestamp prediction. The series includes two all-in-one speech recognition models, Qwen3-ASR-1.7B and Qwen3-ASR-0.6B, which support 52 languages and dialects, as well as a non-autoregressive forced alignment model, Qwen3-ForcedAligner-0.6B, which can align text-speech pairs in 11 languages. The 1.7B version achieves a leading level among open-source ASR models and is competitive with the strongest commercial APIs; the 0.6B version balances accuracy and efficiency, achieving throughput of up to 2000x at a concurrency of 128. The models are based on the audio understanding capabilities of the Qwen3-Omni foundation model, support unified streaming and offline inference, and can transcribe long audio. The project provides a complete inference toolkit, supporting vLLM-based batch inference, asynchronous serving, streaming inference, and timestamp prediction, and can be used quickly by installing the qwen-asr package via PyPI.

Read more