Qwenlm

Beware of fake account

[Qwenlm @Qwenchat]

🔥 Qwen2.5-Max is now ranked #7 in the Chatbot Arena, surpassing DeepSeek V3, o1-mini and Claude-3.5-Sonnet.

It is ranked 1st in math and coding, and 2nd in hard prompts.

👉🏻 Try Qwen2.5-Max here: https://t.co/8b7wwz3voY.…

[Qwenlm @Qwenchat]

Two days ago we released Qwen2.5-Max, and today we update the price of its API:

Input tokens: $1.6 / million tokens
Output tokens: $6.4 / million tokens

We hope this change provides you some help and we'd love to hear your feedback about our new model!

[Qwenlm @Qwenchat]

The burst of DeepSeek V3 has attracted attention from the whole AI community to large-scale MoE models. Concurrently, we have been building Qwen2.5-Max, a large MoE LLM pretrained on massive data and post-trained with curated SFT and RLHF recipes.

[Qwenlm @Qwenchat]

The Future of Technology: Key Trends to Watch in 2025 @Qwenchat

[Qwenlm @Qwenchat]

2 Spatial Understanding
This notebook showcases Qwen2.5-VL's advanced spatial localization abilities, including accurate object detection and specific target grounding within images.See how it integrates visual and linguistic understanding to interpret complex scenes effectively

[Qwenlm @Qwenchat]

1 Computer use

This notebook demonstrates how to use Qwen2.5-VL for computer use. It takes a screenshot of a user's desktop and a query, and then uses the model to interpret the user's query on the screenshot.

https://t.co/Pmkg8H9wyZ

[Alibaba_SpeechAI @TONGYI_SpeechAI]

The Tongyi Speech Team has open-sourced two foundational speech models: SenseVoice and CosyVoice.

😄SenseVoice, a multilingual audio understanding model: Its multilingual speech recognition outperforms Whisper by 50% in Chinese and Cantonese, with inference speed 15 times faster, and it supports state-of-the-art emotion recognition and audio event detection.

😄CosyVoice, a multilingual audio generation model: Trained on over 170,000 hours of multilingual audio data, it supports multilingual speech generation, timbre and emotion control. CosyVoice excels in multilingual speech generation, zero-shot speech generation, cross-lingual voice synthesis, and instruction execution.

We look forward to developers experiencing and using our models, and we appreciate your valuable feedback. If you like our team's open-source projects, please don't hesitate to give us a star on GitHub.

GitHub：
https://t.co/scejYx1ZkO
https://t.co/PTz9xdrrKZ

Demo Samples:
https://t.co/0nPByH1kNS

Techinical Report：
https://t.co/aFMJk2fClu

[Qwenlm @Qwenchat]

Notably, the flagship model Qwen2.5-VL-72B-Instruct achieves competitive performance in a series of benchmarks covering domains and tasks including document and diagrams reading, general visual question answering, college-level math, video understanding, and visual agent #Qwen

[Qwenlm @Qwenchat]

instruction-tuned versions. The flagship model, Qwen2.5-VL-72B-Instruct, is now accessible through the https://t.co/0GgltT79TU
platform, while the entire Qwen2.5-VL series is available on #huggingface and Alibaba’s open-source community Model Scope. https://t.co/wEgnUYAjSB

[Qwenlm @Qwenchat]

Alibaba Cloud Releases Latest AI Models For Enhanced Visual Understanding and Long Context Inputs