​
Intro

Setting Up Ollama for Seamless Integration with NextChat
Please ensure your client version is greater than or equal to v2.11.2 and Ollama version is greater than v0.1.24
Introduction: Ollama has gained popularity for its efficient model management capabilities and local execution. However, due to the current deployment constraints of Ollama and NextChat, some configurations are required to ensure the smooth utilization of Ollama’s model services.
For different system environments, refer to the configuration methods outlined at https://github.com/ollama/ollama/blob/main/docs/faq.md.
1

Configure Ollama CORS:

  • If you are accessing Ollama from the NextChat client, please add the following to your configuration: 
    OLLAMA_ORIGINS=*://localhost
  • If you are using OLLAMA from a non-local source, set the following configuration: 
    OLLAMA_HOST="0.0.0.0"
    It’s recommended to replace “0.0.0.0” with the specific domain or IP address you intend to use.
2

(optional)Configure HTTPS Certificate for Ollama API:

  • This step is necessary only if you are accessing the Ollama API over HTTP from a NextChat instance hosted on an HTTPS website.
3

Config Ollama API in NextChat:

  • Update OpenAI Endpoint to your Ollama deployment endpoint(e.g. http://localhost:11434/)
  • Leave OpenAI API Key empty
  • Set Custom Model to the model you want to use(e.g. gemma, mistral, ‘llama’). Please ensure you have the model installed in your Ollama instance. Settings Page
Note: Ensure that these configurations are set up before attempting to use Ollama’s model services for a seamless experience.