Table of Contents
Chutes has begun rolling out Qwen3.8 Max and DeepSeek V4.1 Pro in beta as the Bittensor Subnet 64 compute platform ends subscription renewals and shifts access to the new models to pay-as-you-go billing.
Existing subscriptions will continue working normally through the end of their current billing periods, then expire automatically. Neither new model is included in those plans, so subscribers who want to use them must access them separately on a usage-based basis.
Chutes attributed the decision to the requirements of the new offerings and conditions in the GPU and compute markets. Subscriptions are “no longer feasible,” the team said in its platform update.
The beta models also have a different security profile from Chutes' confidential-compute offerings. Neither currently supports its trusted execution environment (TEE), confidential-compute, or end-to-end encryption modes. A TEE provides a protected environment for processing workloads. Users who need those protections should weigh that difference when choosing a model.
Chutes provides serverless inference, allowing developers to run AI models without managing the underlying GPU servers themselves. The additions expand the models available through that service, with users paying for consumption instead of receiving access through a recurring subscription.
