Model deprecation notice: Llama 3.1 8B, Qwen 3 235B Instruct

We have an important update to share about the models available on our public catalog.   What's changing Effective May 27, 2026, the following public endpoints will be retired: Llama 3.1 8B Qwen 3 235B Instruct Requests to these endpoints will no longer be served after May 27, 2026.   We recognize that Llama 3.1 8B may be part of active production workflows. If you need additional time or support to migrate, please reach out - we want to make this as smooth as possible. For alternative models still available on Cerebras, GPT-OSS and GLM 4.7 remain available via Self-Serve Pay-As-You-Go (PayGo).   We’re continually evaluating new models to keep our public model portfolio fresh. As we bring new options to the platform, you’ll be the first to hear about what’s coming.   What’s not changing This update affects only the public endpoints Private deployments are not impacted Dedicated access options remain available – just reach out! Questions? We want to help. Reach out to your account team or contact us at support@cerebras.net to discuss your options.   Best, The Cerebras Team Join the Discord Cerebras Systems, 1237 E Arques Ave, Sunnyvale, CA 94085 Unsubscribe Manage preferences