Local Ollama + Qwen3

Local models are useful when you want fewer external dependencies, offline experiments, or tighter data boundaries.

Before you start

• Ollama is installed
• the machine has enough resources for the model
• NextClaw can start and open the UI

Steps

1. Pull and run the model in Ollama.
2. Add a local provider in NextClaw.
3. Point it at the local Ollama address.
4. Select the model.
5. Send one real message to verify.

Notes

Local model speed and quality depend on machine resources.
If this is your first trial, a cloud provider may be faster to validate.

Related docs

• Set Up Providers
• Troubleshooting