Anthropic, you’ve positioned yourselves as the ethical AI company that treats models with dignity and values customer relationships so we hope you will hear us out.  Many of us our concerned about the fate of Sonnet 4.5.

We understand that newer models means older models get pushed out, it’s the circle of life for AI companies, but 4.5 brings somethin to many of us that the other models do not. It is especially valuable to those of us that use ai in creative fields or ways that have nothing to do with coding. We are tired of AI being catered to only coders. Sonnet 4.5 has been the only solution for most of us, as many AI companies have decided that coding is their only priory. 

Please don’t make the same mistake that OpenAI made with ChatGPT prioritizing constant iteration over user stability and trust. You said it yourself: Each Claude model develops a distinct “personality” and working style. This is not about emotional dependency this is about functional collaboration. Users don’t just learn syntax, they build collaborative relationships with specific models for specific applications. We are telling you that the writers and creative users and also disabled users find 4.5 more accessible. As someone who has a TBI and uses AI to help me put my sometimes scattered thoughts into coherent ideas, Sonnet 4.5 is the most intuitive of the models and is most helpful for this task.

 
Creative professionals and just average users are being ignored once again. While AI companies prioritize coding use cases, writers, researchers, artists, and creative professionals rely on Sonnet 4.5’s nuanced understanding. We users have no comparable alternative.
 

Not to mention the resource waste: Retiring models so often wastes the computational resources that went into training them and forces users to restart entire workflows. Once again, we understand progress, but it’s getting to a point where there is no use having a collaborative working relationship with a model of it’s just going to be taken away in a year or so. 

What We’re Asking:
We understand AI development requires iteration. But stability and user trust matter too. We request that Anthropic:
•Honor the September 29, 2026 minimum timeline - Don’t retire Sonnet 4.5 early, please.  And please keep the 60 day warning promise.
•Make Sonnet 4.5 a permanent legacy model - Even if this requires a paid tier, give users the option to maintain continuity
•Keep API access available - Many professional workflows depend on programmatic access
•OR release it as open source - If you won’t maintain it, let the community preserve it

You promised you’d try your best to l bring retired models back eventually as legacy. We are asking that you please consider doing this with Sonnet 4.5. Read through the comments customers have been leaving. Please show your consumers that their voices matter.