AI 工具情报

探索 AI 工具

按工作流、定价、市场表现及背后的生态主体,追踪实用的 AI 产品。

工具顾问

描述您的工作流。

基于已发布的 AI 工具资料排序。

顾问

描述工作流,而非关键词。

顾问会将“为营销活动缩略图生成图像”这类需求转化为带有匹配理由的工具推荐排序。

工具索引

显示 6 款工具

SEA-LION

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.

免费 / SEA-LION / LLM / Open source / Multilingual

Falcon LLM

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.

免费 / Technology Innovation Institute / LLM / Open source / Arabic

Maritaca AI (Sabiá)

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.

免费增值 / Maritaca AI / LLM / API / Portuguese

Upstage Solar API and Document Parse

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.

付费 / LLM / Document AI / API

DeepSeek API

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.

付费 / DeepSeek / LLM / API

Cohere API

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.

免费增值 / LLM / API / Embeddings