Request to Continue Access to Llama 4 Maverick Model

Feature Requests
1

I am writing to discuss the recent change regarding the availability of the Llama 4 Maverick model(depreciated on march 9th 2026). For our use case, this model has been extremely important because it provides strong support for multiple Indian languages. Our application relies heavily on multilingual understanding, particularly for Indian languages, and Llama 4 Maverick has been performing significantly better for this requirement.

Since the switch to GPT OSS, we have observed a noticeable degradation in our system’s performance. In comparison, GPT OSS is currently not delivering the same level of accuracy or language support that we previously experienced with Llama 4 Maverick.

Given the importance of multilingual capability in our system, we would like to check if there is any way for us to continue accessing the Llama 4 Maverick model, or if there are any alternative options available that would allow us to maintain similar performance.

We would greatly appreciate any guidance or solutions you can provide.

Thank you for your support, and we look forward to your response.

2

This would be beneficial to us also. Our LLM prompt worked so well on maveric, but is completely broken on gptoss120b, scout, kimi etc. I have rebuild a new prompt which works but not nearly as well as maveric

3

Yes there is no real substitute for Maverick but given the whole situation with no new model being added and lack of communications I preferred to switch using maverick from Openrouter. If it ever comes back to Groq I’ll be happy to come back

4

Not a bad Idea, I think I should do the same. Openrouter here I come :slight_smile: