Google Gemini 3 launch November 2025: What’s live on day one
The Google Gemini 3 launch November 2025 hit Search’s AI Mode on day one, according to the Wall Street Journal and InfoQ. Additionally, it landed across the Gemini app, AI Studio, Vertex AI, the Gemini CLI, and the google antigravity ide. Together, that delivery widened access immediately.
In practice, gemini 3 search integration puts model answers directly into Search’s AI experiences. Moreover, it sets distribution before long-tail developer rollouts.
Rollout timeline for the Google Gemini 3 launch November 2025
On Nov 18, Google launched the first Gemini 3 series model as the gemini 3 pro preview. The release notes record, “Launched the first Gemini 3 series model, gemini-3-pro-preview,” along with new “thinking levels” and “thought signatures.”
On Nov 19, the alphabet stock reaction was immediate. Alphabet shares rose as much as 6.9% and set a record, Bloomberg reported.
On Nov 20, Google added the Pro Image pathway. The notes state, “Released Gemini 3 Pro Image Preview, gemini-3-pro-image-preview,” for image-generation workflows.
Here’s the twist: Reasoning and multimodal advances outpacing prior versions
Reports say Gemini 3 improves planning, tool use, and long-horizon tasks versus earlier releases. The Wall Street Journal says the model surpassed rivals on multiple benchmarks and credits leadership focus. Consequently, expectations for enterprise-grade reasoning rose.
InfoQ notes Gemini 3 Pro accepts text, images, video, audio, and PDFs. It supports input windows up to 1,048,576 tokens and outputs up to 65,536 tokens. Therefore, that multimodal scope is built for complex, multi-step flows.
Deep Think: Higher-intensity reasoning mode with safety gating
Google introduced the deep think reasoning mode alongside Gemini 3 Pro. It targets the hardest reasoning workloads and long-horizon planning, InfoQ reports.
Access begins with safety gating. Additionally, broader access to the deep think reasoning mode is expected after more testing. As one report put it, “Deep Think is next level.”
Developers: Antigravity, previews, and new controls
Developers can reach Gemini 3 through the API and an agent-first environment. The google antigravity ide centers on multi-agent, tool-using workflows with editor, terminal, and browser access, according to InfoQ.
The gemini 3 pro preview is the entry point for server calls today. Additionally, Google’s “thinking levels” and “thought signatures” expose tunable reasoning behavior in the API. As a result, the google antigravity ide can orchestrate multi-step plans end to end.
Market impact: Alphabet shares hit records on Gemini 3 enthusiasm
The alphabet stock reaction was strong. Shares rose as much as 6.9% on Nov 19 with heightened call activity, Bloomberg reported.
Investors are betting on Search tie-ins and enterprise adoption. However, the alphabet stock reaction will depend on sustained usage and economics.
What’s next: Wider Deep Think access and continuing previews
Expect ongoing preview updates across API and image-generation endpoints. Google’s notes show a steady rollout cadence across Gemini 3.
Additionally, broader availability of the deep think reasoning mode should follow safety reviews. And as gemini 3 search integration matures, partners will watch how the Google Gemini 3 launch November 2025 translates into durable user behavior.
