Google Gemini June 2026: Interactions API GA, Gemini Omni Video, and Gemini 3.1 Updates

Google Gemini June 2026: Interactions API GA, Gemini Omni Video, and Gemini 3.1 Updates Google's June 2026 Gemini announcements center on the Interactions API reaching general availability, Gemini Omni for video generation, and expanded multilingual support in Google Sheets. The Interactions API is now the primary interface for building stateful, agentic applications with Gemini. Interactions API: General Availability After a public beta launched in December 2025, the Gemini Interactions API has reached GA as of June 22, 2026. The API is now the default for Google AI Studio, the Gemini API, and all documentation. Key GA features: Managed Agents: A single API call provisions a remote Linux sandbox where an agent can reason, execute code, browse the web, and manage files. The Antigravity agent ships as the default. Custom agents can be defined with instructions, skills, and data sources. Background execution: Set on any call. The server runs the interaction asynchronously, with results retrieved when ready. Tool improvements: Mix built-in tools (Google Search, Google Maps) with custom functions in a single request. Tool results can now return images alongside text. Deep Research upgrades: Two new agent versions—speed vs. depth—plus collaborative planning, native charts and infographics, and multimodal grounding with images, PDFs, and audio. Media generation: Image generation with Nano Banana 2 and Google Image Search grounding, music with Lyria 3, and expressive multi-speaker TTS. Schema change: The legacy role-based schema (, ) is replaced by a steps-based schema where every action (user_input, thought, function_call, model_output) is its own typed step. Cost optimization: Flex and Priority tiers let developers optimize for cost or latency. Flex offers 50% cost reduction compared to standard pricing. The legacy API remains fully supported and will continue receiving new mainline models, but frontier capabilities for long-running agents will land exclusively on the Interactions API. Gemini Omni: Video Generation Gemini Omni rolled out in May 2026 to Google AI subscribers 18+, letting users create video content through natural language prompts. Omni is positioned as a creative partner for video creation. The system generates video clips from text descriptions, with style and content controlled through prompt engineering. Quality and length limits are tied to subscription tier. Gemini 3.1 Pro Gemini 3.1 Pro launched in February 2026 as Google's flagship model for complex tasks, now available globally in the Gemini app with higher limits for Google AI Pro and Ultra subscribers. The 3.1 Pro release emphasized: Improved reasoning across multi-step tasks Better context window utilization (up to 2M tokens) Reduced hallucination rates on factual queries Native tool use improvements Sheets: 28 Languages Google expanded Gemini's spreadsheet capabilities in Sheets to support 28 languages, letting global teams build models, charts, and formulas using natural language prompts in their native language. Supported languages include Spanish, Portuguese, Japanese, Korean, Italian, French, German, Chinese Simplified, Dutch, Hebrew, Polish, Turkish, Czech, Indonesian, Malay, Swedish, Danish, Norwegian, Arabic, Finnish, Vietnamese, Ukrainian, Greek, Thai, Romanian, Russian, Catalan, and Hungarian. Promotional higher limits run through July 15, 2026, after which per-user quotas apply. The feature requires Gemini for Workspace enabled plus smart features turned on. $100 Google AI Ultra Plan Google introduced a $100 USD Google AI Ultra plan in May 2026, sitting above the existing Pro tier. The Ultra plan provides: Highest usage limits for Gemini API and AI Studio Priority access to new model releases Increased context windows Early access to experimental features Developer Implications The Interactions API GA is the most significant developer-facing change. The shift from role-based to step-based schema requires migration for existing projects, but Google has published a migration guide. For developers building agentic applications, Managed Agents simplify the infrastructure—Google handles the sandbox provisioning. For organizations with compliance requirements around data processing, the managed agent model means understanding where agent execution happens and what data it touches. The 50% cost reduction on the Flex tier makes high-volume agentic applications more economically viable. For use cases that previously required careful token budgeting, Flex tier pricing changes the calculus.