Ollama launches web search API to boost AI model accuracy and reduce hallucinations

Ollama launches web search API to boost AI model accuracy and reduce hallucinations

Ollama has introduced a new web search API, expanding the platform’s capabilities to provide real-time information for open AI models. This release allows models — including gpt-oss — to access current web data, supporting long-running research tasks and helping to reduce hallucinations while improving accuracy for users who rely on up-to-date content.

Alongside the main API, web search functionality is available as a REST endpoint and includes advanced integrations through Ollama’s Python and JavaScript libraries. Developers can also activate web search in any MCP client by using the Python MCP server, making workflow integration more straightforward across technical stacks.

To get started, users can simply generate an API key from their Ollama account. Web search is included for free with all standard accounts, while those requiring higher rate limits can upgrade their Ollama subscription for expanded access.

by Paul

VoidyWay
VoidyWay found this interesting
Ollama iconOllama
  114
  • ...

Ollama enables local deployment of language models like Llama 3 and Code Llama, allowing for customization and offline AI development. It's designed for creating personalized AI chatbots and writing tools. Rated 4.6, Ollama is AI-powered, ad-free, and functional offline. Key alternatives include ChatGPT, Claude, and Bard.

No comments so far, maybe you want to be first?
Gu