it never works for me, always errors.

[veerababumanyam]

Which version of the app are you using?

v3.3.22

Which API Provider are you using?

OpenRouter

Which Model are you using?

tried DeepSeek, Ollama

What happened?

Error
Unexpected API Response: The language model did not provide any assistant messages. This may indicate an issue with the API or the model's output.

Error
Unexpected API Response: The language model did not provide any assistant messages. This may indicate an issue with the API or the model's output.

Roo is having trouble...
Roo Code uses complex prompts and iterative task execution that may be challenging for less capable models. For best results, it's recommended to use Claude 3.5 Sonnet for its advanced agentic coding capabilities.

Steps to reproduce

1. Just install and use any model from ollama or deepseek

Relevant API REQUEST output

Unexpected API Response: The language model did not provide any assistant messages. This may indicate an issue with the API or the model's output.

Additional context

No response

[samhvw8]

can you export chat log ?

[RedCloudDC]

This issue remains despite all the recent updates to roo-code it's not a good experience when using Ollama local models (especially small 8b or 10b models) it throws an error recommending Sonnett-3.5 which defeats the purpose of Ollama - can you get Ollama working better with Roo-Code?

[Nick2121]

I'm having the same exact issue and have tried numerous models. I think I've narrowed it down to the models not using the correct format for tool use calls back to roo code. It would be easier if there was a system prompt that roo code could provide that would show the proper tool use call format. That has been difficult to find but based on when the issues arise it seems to be the cause.

Also, for the love of all that is Holy quit telling me to to use Claude 3.5, of course that works, it's also the most expensive at this point and the fact I'm using roo-code instead of paying for another service should tell you we're trying to use local hardware and models to do this.