Nov 3, 2026: Cortex AI Gateway inference, observability, and cost management (General availability)¶

Inference, observability, and cost management for Cortex AI Gateway are now generally available. The gateway gives AI clients in your organization one governed endpoint to reach models, records a trace of every request, and attributes spend to the users who made it.

The following capabilities are now generally available:

• Inference: send requests to the gateway endpoint in the OpenAI Chat Completions or Anthropic Messages API format, from the official SDKs or a coding agent. See Inference with Cortex AI Gateway.
• Observability: query the spans for each request, including the models called, timing, token counts, and errors, with AGENT_TRACE_TABLE. Client applications can also export their own OpenTelemetry traces to the gateway. See Observability for Cortex AI Gateway.
• Cost management: report on spend with the AI_GATEWAY_USAGE_HISTORY view, track spend by team with shared resource budgets, and block individual users at a limit with per-user quotas. See Cost management for Cortex AI Gateway.

This release also adds trace analytics, in preview. On the Analytics tab of the gateway’s Observability page in Snowsight, you can chart model usage, request failures and latency, MCP tool calls, and conversation length from the trace table without writing SQL. See Trace analytics for Cortex AI Gateway.

Tools for Cortex AI Gateway remains in preview.

To get started, sign in to Snowsight and select AI Gateway in the navigation menu. For more information, see Cortex AI Gateway.