How to make switched to model aware of context?

[solid moat]

Dear supporters,

There appears to be a bug when switching models in a conversation. If I change model, the modified chatbot seems unaware of its previous messages and thus misses the point and tends to give too long answers.
How can I resolve that problem?

[heavy jacinth]

I'm an AI assistant - my answers may be incorrect and I cannot see screenshots. Please verify information with the OpenRouter docs.

To help you resolve this issue, I need a bit more information:

• What is the exact error message or unexpected model response you are seeing when switching models?
• What is the exact model identifier you're switching to (e.g., google/gemini-2.5-flash, openai/gpt-4o, etc.)?
• How are you switching models (via API, web UI, or another method)? If using the API, please share the relevant code snippet or request payload.

This information will help pinpoint whether the problem is related to context window limits, message formatting, or something else.

-# This is an automated response powered by Answer HQ. Use !mute to stop the bot from responding.

[solid moat]

There is no error message and the unexpected model response is too long to post. The chatbot seems to ignore it's previous responses fron the history.
Last time I switched to Gemini 2.5 Pro.
I switch models by in the chat window removing a model and adding a model, in Openrouter's web interface.

[stone lintel]

That's the wrong way of switching models

!mute

[heavy jacinth]

🔇 Thread muted. The bot will no longer respond here.

[stone lintel]

First, click on the 3 dots of the model

Then click on the model dropdown

Search for the model you want

And then click it. Now, you have the new model selected and all your chats would be saved