تحليلات أدوات الذكاء الاصطناعي
تابع منتجات الذكاء الاصطناعي المفيدة حسب سير العمل والتسعير والانتشار والجهات التي تقف وراءها في المنظومة.
المستشار
اطلب سير عمل، لا كلمة مفتاحية.
يحوّل المستشار متطلبًا مثل «توليد صور مصغّرة للحملات» إلى اقتراحات أدوات مرتبة مع أسباب ملاءمتها.
Family of open-source multilingual and multimodal language models built by AI Singapore for Southeast Asian languages and cultural context. The models can be downloaded from Hugging Face and Ollama for research or commercial use, and an API key is available through the SEA-LION playground.
Family of open or open-access language models from Abu Dhabi's Technology Innovation Institute, including hybrid, reasoning, multimodal and Arabic-focused variants. Weights for many models are published on Hugging Face, and licence terms differ by model, with some under or based on Apache 2.0.
Brazilian developer of the Sabiá language models, tuned for Portuguese and Brazilian context and offered through a chat app and an API that works with the OpenAI client library. The chat has a free tier capped at 20 messages a week; the API is billed per token in reais, with an option to keep inference in Brazil.
Korean AI company offering Solar language models and Document Parse, which converts PDFs and scanned documents into structured text, through a developer API. Usage is billed per token or per page, and Studio agents include 10 free runs.
Pay-per-token API access to DeepSeek's large language models, currently a Flash model and V4 Pro. Rates are halved outside weekday peak hours, and input served from cache costs far less than uncached input.
Developer platform offering Cohere's Command language models plus Embed and Rerank models for search and retrieval, alongside the North enterprise agent platform. New accounts get a free rate-limited trial key for non-production use; production usage is billed per token or per hosted instance.
