PLATFORMS Signal 673 2 feeds carried it
Hy4 Preview now available on AI Gateway
↗
Engineers can now call Tencent's open-source 770B-parameter Hy4 Preview Mixture-of-Experts model via Vercel AI Gateway using the tencent/hy4-preview identifier.
AI Gateway provides a unified API for calling models, tracking usage and cost, and configuring retries, failover, and performance optimizations for higher-than-provider uptime. It reflects provider pricing with no markup and does not charge a platform fee on inference, including BYOK requests. Engineers can therefore use the model for long-horizon coding, document analysis, game development, and scientific reasoning while benefiting from gateway-level observability and reliability.
Written by elseif from the cluster below · every claim links back to a sourceThe three things worth knowing
Hy4 Preview is an open-source Mixture-of-Experts model with 770B total parameters, 49B active per token, and a 1M-token context window.
Access is via the model identifier tencent/hy4-preview in Vercel AI Gateway, with example code provided for the AI SDK and coding-agent setup.
AI Gateway delivers unified API calls, usage tracking, cost reflection, retries, failover, Zero Data Retention support, budgets for API keys, routing rules, and no platform fee on inference.
THE READ
What the cluster adds up to.
Hy4 Preview, an open-source Mixture-of-Experts model from Tencent, has been added to Vercel’s AI Gateway. Engineers can now select the model using the identifier tencent/hy4-preview in the AI SDK or via a coding-agent setup. The AI Gateway provides a unified API for calling models, tracking usage and cost, and configuring retries, failover, and performance optimizations. This gives access to a model with 770B total parameters, 49B active per token, and a 1M-token context window.
Adopting the model does not incur a platform fee; AI Gateway reflects the provider’s pricing with no markup. To use it, developers set the model field to tencent/hy4-preview and can follow the provided code snippet for streamText. For coding agents, the vercel ai-gateway coding-agents setup command connects agents such as Claude Code, Codex, OpenCode, Cursor, Pi, and more, after which the model can be selected inside the agent. API keys can be budgeted, usage reported, and Zero Data Retention enabled through the gateway’s built-in features.
The model is positioned for long-horizon coding, document analysis, game development, and scientific reasoning tasks that benefit from its large context window. AI Gateway provides higher-than-provider uptime through retries, failover, and performance optimizations. The gateway also includes built-in custom reporting, Zero Data Retention support, budgets for API keys, and routing rules. These features help operators monitor cost and reliability without changing the underlying model pricing.
Written by elseif from the cluster below · checked for specifics the sources never containedTHE CLUSTER