Mistral Vibe: Google Vertex AI Support for Enterprise Cloud Deployments

Google Vertex AI Support Arrives in Mistral Vibe 2.2.0

Released on February 17, 2026, Mistral Vibe 2.2.0 adds native integration with Google Vertex AI, marking a significant expansion of the coding agent's cloud deployment options.

What This Means for Developers

Prior to this release, Mistral Vibe primarily connected through Mistral's own API endpoints. With Google Vertex AI support, enterprise teams can now route Mistral Vibe traffic through Google Cloud's managed infrastructure. This is particularly relevant for organizations that have existing Google Cloud commitments, consolidated billing requirements, or compliance policies that mandate specific cloud providers.

Google Cloud has offered Mistral models on Vertex AI since mid-2024, including Codestral and Mistral Large variants. The Vibe CLI integration means developers can now use these cloud-hosted models directly from their terminal-based coding workflows.

Additional Features in 2.2.0

The same release includes several other notable additions:

• Telemetry System: User interaction and tool usage events can now be sent to a datalake, configurable via the disable_telemetry setting for teams that need usage analytics.
• Agent Skills Standard Compatibility: Skill discovery now supports .agents/skills/ directories in addition to .vibe/skills/, aligning with the broader Agent Skills specification.
• Proxy Wizard: A proof-of-concept proxy setup tool for both CLI and ACP modes, helping developers behind corporate proxies configure their connections.

Enterprise Context

The Vertex AI integration fits into Mistral's broader enterprise strategy. With on-premises deployment already available through open Devstral 2 weights, and Mistral's own API as the primary cloud option, Google Vertex AI provides a third deployment path that leverages an established cloud provider's security, compliance, and billing infrastructure.