Skip to main content
MA

Maritaca AI

Maritaca AI offers Portuguese-optimized Large Language Models (LLMs) via API for developers. Their specialized models deliver superior performance and cost-efficiency for advanced natural language processing tasks in Portuguese.

Campinas, BrazilFounded 20221710K+ followers
Updated 4 months ago

Funding

Funding not disclosed

Funding rounds are not available yet.

Founders

Product

Problem

Developing applications that require sophisticated natural language processing in Portuguese often necessitates the use of general-purpose large language models (LLMs) that may not be optimized for the nuances of the language. This can lead to suboptimal performance, increased operational costs, and a lack of specialized contextual understanding.

Solution

Maritaca AI provides a suite of Portuguese-centric Large Language Models (LLMs) accessible via API, designed to deliver superior performance and cost-efficiency for developers. Their models, such as Sabiá-3.1 and Sabiazinho-3, are specifically trained on Portuguese language data, enabling detailed explanations, contextually relevant responses, and advanced text generation capabilities. By leveraging specialized training methodologies and efficient token processing, Maritaca AI empowers developers to integrate high-quality AI functionalities into their applications at a significantly reduced cost compared to generic LLM providers. The platform offers flexible pay-as-you-go pricing based on token consumption, making advanced AI accessible for a wide range of use cases.

Target Audience

The primary target audience includes software developers and technology companies building applications that require advanced natural language processing capabilities specifically for the Portuguese language.

Features

  • Portuguese-optimized LLMs (Sabiá-3.1, Sabiazinho-3) with varying context window sizes (up to 128,000 tokens for Sabiá-3.1).
  • API access with pay-per-token pricing for both input and output tokens.
  • High throughput processing, with models capable of handling 500,000 input tokens per minute.
  • Demonstrated performance on national Portuguese language examinations, indicating strong linguistic comprehension and generation capabilities.
  • Support for detailed explanations, contextual responses, and creative text generation.
  • Competitive pricing, advertised as up to 4x lower cost per token compared to competitors.
  • Initial credit offering of R$20 for new API users.
This profile is AI-generated and may contain inaccuracies.