Todd

Claude Fable 5.2 is now available on Zevio at 45% less. 1M context. Text, image, and file inputs. More room for your biggest tasks. Input: $4.40 / 1M tokens Output: $22.00 / 1M tokens Model ID: anthropic/claude-fable-5.2 Use it in your workflow or try it on Zevio zevio.sh Discounted AI. Private billing.

We’ve been quietly working on something that changes the speed profile of Zevio. A new inference provider is coming, bringing throughput of up to 700 tokens per second across selected models and a much faster experience throughout the Zevio ecosystem. We’re not revealing the name just yet. But if fast AI matters to you, this is one to watch. More soon. ethereum:0x5df882f8788071b67d0283f958ca3cdc17f1ae2b

GLM 5.3 Flash Uncensored is now available on Zevio at 45% less. Less filtering. More freedom to explore, create, and build. Per 1M tokens: • Input: $0.066 • Cached input: $0.0165 • Output: $0.275 More possibilities. Less cost. Try it now: zevio.sh Discounted AI. Private billing.

GLM and Qwen Abliterated are shipping on Zevio today. Over the past few days, usage across Zevio has continued to grow, and that means the infrastructure behind the models has to scale with it. We’ve just finalized a new provider arrangement to bring higher-throughput LLM capacity into Zevio, with significantly higher TPS so responses can be served faster and larger workloads can move through the platform more efficiently. The first models coming through this expanded capacity will be GLM Abliterated and Qwen Abliterated, giving users more high-performance, less restricted options across Workspace, Harness, and the Zevio API. This isn’t only about adding more models. As usage grows, we’re continuously improving the infrastructure underneath Zevio so the experience gets faster, more reliable, and better prepared for heavier workloads. You can watch Zevio usage grow in real time here: zevio.sh/usage GLM Abliterated and Qwen Abliterated go live today. More capacity. Faster inference. More choice.