Gemini models are good for this! There’s also a new model on OpenRouter that you can try w/ a 2M token context window… should be more than enough

→ View original post on X — @mattshumer_