Google has shut down three Veo video generation models in the Gemini API effective June 30, 2026, forcing immediate migration for any remaining production workloads.
The deprecated models — veo-2.0-generate-001, veo-3.0-generate-001, and veo-3.0-fast-generate-001 — are no longer available. Developers must switch to Veo 3.1 preview models (veo-3.1-generate-preview or veo-3.1-fast-generate-preview) via the standard Gemini API or to the generally available Veo 3.1 versions through the Gemini Enterprise Agent Platform to avoid service interruptions.
Deprecation Data
**Effective date**: June 30, 2026 (immediate shutdown)
**Affected models**: veo-2.0-generate-001, veo-3.0-generate-001, veo-3.0-fast-generate-001
Migration targets:
- Preview: veo-3.1-generate-preview, veo-3.1-fast-generate-preview (Gemini API)
- GA: Available via Gemini Enterprise Agent Platform’
Key upgrades in Veo 3.1: Native synchronized audio generation (dialogue, sound effects, ambient audio at 48kHz stereo), video extension capability, improved image-to-video with better prompt adherence and character consistency, native support for portrait (9:16) and landscape (16:9) formats, and higher-fidelity output up to 4K
Impact: Calls to deprecated model IDs will fail; no extended grace period noted in the changelog
The move comes as Google continues rapid iteration in generative video. Veo 3.1, first introduced in preview in October 2025, adds native audio as a core capability — a meaningful step beyond prior versions that required separate audio layering. It also introduces practical production features such as extending existing Veo-generated clips and generating from specified first and last frames.
Google has grouped this shutdown with other model cleanups, including the planned August 17, 2026 deprecation of certain Imagen 4 and Gemini 3 image models.
The pattern underscores Google’s approach of sunsetting older versions to consolidate usage on its latest frontier models.
For high-intent readers tracking generative AI infrastructure, today’s cutoff creates near-term operational pressure for any teams still running on Veo 2.0 or 3.0 in production. At the same time, the newer models deliver tangible capability lifts — particularly native audio and extension tools — that can reduce downstream editing costs and improve output quality for marketing, entertainment, and enterprise video workflows.
The forced migration highlights both the velocity of model advancement in video generation and the execution risk for developers who lag on API updates.