12 articles
Explore text, image and audio models, comparing supported modalities and deployment requirements.
Check exact input and output modalities: understanding images does not imply generating them. Review video sampling, audio latency, context limits and inference-framework support for your task.
MOSS-VL-Realtime technical introduction and quickstart: OpenMOSS has open-sourced an 11B realtime video understanding model with a 256K context window, timestamped frame input, streaming interaction, proactive silence, and dynamic correction.
Google releases Gemma 4 multimodal models with dense and MoE architectures, 256K token context, native reasoning mode, and deployment from mobile to server. Four model sizes available under Apache 2.0.
Qwen3.5-Omni: Native multimodal model with 256k context, 10-hour audio, 4M frame video support. Thinker-Talker + Hybrid-Attention MoE architecture achieves SOTA in 215 benchmarks.
Qwen3.5 MoE series (0.8B-397B) with Gated DeltaNet hybrid attention. 9B outperforms 120B models, 0.8B processes video on smartphones. 1M context, native multimodal, 201 languages.
GLM-4.6V free open-source multimodal AI model guide. 106B/9B versions, 128K context, image recognition, OCR, PDF understanding, video analysis.
Qwen3-Omni-modal model for text, image, audio and video with real-time speech. Thinker–Talker + MoE, multi-codebook for low latency; 119 languages; vLLM/Transformers tips.
Meta Llama 4 faces ranking manipulation allegations while dealing with executive departures, putting its AI strategy to the test
Comparing Qwen2.5-Omni and MiniCPM-V (2.8B) local deployment, Analysis of parameters, hardware needs, and optimal use cases.
xAI's Grok 3 launches powerful image editing via Aurora model, challenging Google's Gemini. With chat-driven editing and superior character consistency, Grok 3 takes lead while Gemini remains in testing.
CogAgent-9B: A 9B-parameter GUI agent by Zhipu AI and Tsinghua University that excels in interface understanding and automation, outperforming other models in MM-Vet and more benchmarks
Discover Google Gemini 2.0 Flash: 2x faster performance, enhanced multimodal capabilities, and breakthrough AI features. Learn how this next-gen AI model transforms development and analysis.
Ivy-VL: 3B lightweight vision-language model outperforms 7B models, enables real-time AI glasses, ranks #1 on OpenCompass under 4B. Open-source edge AI solution by AI Safeguard, CMU & Stanford.