Politics

Google's Gemini 3 Deep Think enhances reasoning for high-end AI subscribers

Monday, 8 December 2025 8:17PM UTC

Google announces Gemini 3 Deep Think, a cutting-edge reasoning mode within its Gemini app, offering advanced multi-step problem-solving capabilities to premium subscribers and targeting enterprise and research applications.

Google has rolled out Gemini 3 Deep Think, an enhanced reasoning mode within its Gemini app intended for demanding problem‑solving in mathematics, science and logic. According to the original report, Google describes it as the model’s "strongest reasoning capabilities yet," a feature now available to subscribers of the premium Google AI Ultra plan. ^[1]^[6]^[2]

Deep Think prioritises methodical, multi‑step deliberation over rapid answers: it breaks problems into components, explores parallel hypotheses and applies self‑verification to refine conclusions. Industry data shows the mode builds on the recently introduced Gemini 3 architecture and its multimodal capabilities. ^[1]^[6]^[2]

Benchmark results cited by Google and independent outlets indicate substantial gains: Deep Think scores 41.0% on Humanity’s Last Exam and 45.1% on the ARC‑AGI‑2 benchmark when allowed code execution, outperforming earlier Gemini variants and several competitors on abstract reasoning tests. Publications and technical analyses also highlight improvements on domain‑specific evaluations such as GPQA Diamond. ^[3]^[5]^[7]

Technically, Gemini 3 Deep Think leverages Gemini 3’s expanded context and multimodal processing , including a reported 1 million‑token context window , enabling it to synthesise large amounts of text, images, video and code for complex, iterative tasks. The model supports extended outputs (up to tens of thousands of tokens) and tool usage that benefits simulation, debugging and research workflows. ^[1]^[5]^[6]

Access is currently gated: Deep Think is labelled experimental in the Gemini app and reserved for Google AI Ultra subscribers, a tier reported to cost around $250 per month for individual users. That positioning targets enterprises, researchers and high‑end individual users rather than the general consumer market. ^[1]^[2]^[3]

Early user feedback and social posts from developers suggest notable benefits for debugging large codebases, modelling scientific scenarios and running multi‑step reasoning tasks, while also expressing frustration about limited access and calls for tiered availability or trials. According to the original report, quota increases for Ultra subscribers aim to let power users experiment without frequent interruptions. ^[1]^[2]^[3]

Google has emphasised safety testing prior to release, yet industry observers warn of persistent challenges such as bias in hypothesis evaluation and the limits of automated self‑verification. The company’s staged rollout and experimental label reflect a balance between capability deployment and ongoing evaluation. ^[1]^[6]^[3]

Looking ahead, Google signals that elements of Deep Think may be iterated and potentially broadened to other tiers or products over time, with further integration across Google Cloud, Vertex AI and agentic platforms like Antigravity , moves that would strengthen enterprise adoption and developer workflows. Market watchers also note potential investor interest as premium AI subscriptions become a monetisation avenue. ^[6]^[4]^[5]

📌 Reference Map:

##Reference Map:

^[1] (WebProNews) - Paragraph 1, Paragraph 2, Paragraph 5, Paragraph 6, Paragraph 7
^[2] (Android Central - Deep Think availability) - Paragraph 1, Paragraph 3, Paragraph 6
^[3] (LiveMint) - Paragraph 3, Paragraph 6, Paragraph 7
^[4] (Android Central - Gemini 3 overview) - Paragraph 8
^[5] (Technology.org) - Paragraph 3, Paragraph 4, Paragraph 8
^[6] (Google Blog) - Paragraph 1, Paragraph 2, Paragraph 4, Paragraph 7, Paragraph 8
^[7] (InfotechKeeda) - Paragraph 3, Paragraph 4

Source: Noah Wire Services

More on this

https://www.webpronews.com/google-launches-gemini-3-deep-think-premium-ai-for-advanced-reasoning/ - Please view link - unable to able to access data
https://www.androidcentral.com/apps-software/ai/gemini-3-deep-think-is-now-available-in-the-gemini-app-but-only-for-subscribers - Google has officially released Gemini 3 Deep Think, the most advanced AI reasoning model in its Gemini 3 series. Initially introduced alongside Gemini 3 Pro, which now leads performance benchmarks like LMArena and WebDev Arena, Deep Think is designed specifically for complex math, science, and logic challenges. The model excels in advanced reasoning tasks using parallel hypothesis exploration and has outperformed Gemini 3 Pro in key benchmarks, scoring 41% on Humanity’s Last Exam and 45.1% on ARC-AGI-2 with code execution. Deep Think is currently available in the Gemini app but is exclusive to Google AI Ultra subscribers, who pay $250 per month. It is marked as 'experimental' and can be accessed through the Tools menu under the Thinking model setting. The model replaces Gemini 2.5 Deep Think and is positioned for power users and business applications. Regular Gemini or AI Pro users do not currently have access to this advanced feature.
https://www.livemint.com/technology/tech-news/google-rolls-out-gemini-3-deep-think-to-deliver-advanced-reasoning-how-powerful-is-it-and-who-can-use-it/amp-11764909501711.html - Google's Gemini 3 Deep Think mode offers a substantial upgrade over previous versions, enabling the system to process multiple lines of reasoning in parallel. This improves its ability to handle problems that typically challenge even the strongest AI models. The company highlighted that Gemini 3 Deep Think performs exceptionally well on some of the toughest evaluation benchmarks. It reaches 41.0% on Humanity’s Last Exam without tools and an unprecedented 45.1% on ARC-AGI-2 with code execution, setting a new standard for AI-driven reasoning. These advances build on the achievements of the Gemini 2.5 Deep Think variants, which recently matched gold-medal performance levels in competitions such as the International Mathematical Olympiad and the ICPC World Finals. Ultra subscribers can start using the new feature immediately. Deep Think mode can be activated directly from the Gemini app by selecting 'Deep Think' in the prompt bar and choosing Gemini 3 Pro from the model options.
https://www.androidcentral.com/apps-software/ai/gemini-3-deep-think-googles-most-advanced-reasoning-feature-now-available - Google has unveiled Gemini 3, its most advanced AI model to date, featuring the Gemini 3 Pro variant now available in Search, the Gemini app, and developer tools like Google AI Studio and Vertex AI. Offering leading performance in reasoning and multimodal understanding, Gemini 3 Pro tops key AI benchmarks such as LMArena and WebDev Arena, outperforming its predecessor Gemini 2.5 Pro and other competitors like xAI and Anthropic. Users can utilize the model in AI Mode within Google Search and benefit from smart, concise, and direct answers for complex tasks. Gemini 3 also introduces Gemini Agents capable of handling multi-step tasks, while a more advanced iteration, Gemini 3 Deep Think, is in preview for Google AI Ultra subscribers. As the backbone for Google’s new Antigravity development platform, Gemini 3 enables autonomous AI agents with editor, terminal, and browser access for streamlined coding. Broader rollout plans include automatic selection of models in Search for optimal task efficiency, with features like generative UI and real-time simulations enhancing user interactions across platforms.
https://www.technology.org/2025/11/19/google-gemini-3-released-ai-now-powers-every-search-query/ - Gemini 3 Deep Think extends the model’s reasoning and multimodal understanding to handle more complex challenges. Testing shows Deep Think outperforms Gemini 3 Pro on Humanity’s Last Exam (41.0% without tools) and GPQA Diamond (93.8%). The mode achieves 45.1% on ARC-AGI-2 with code execution, demonstrating its ability to solve novel problems. Gemini 3 synthesizes information across text, images, video, audio, and code. The model combines advanced reasoning, vision, spatial understanding, multilingual performance, and a 1 million-token context window. Users can feed Gemini 3 handwritten recipes in different languages to create translated family cookbooks. The model processes academic papers, video lectures, or tutorials to generate interactive flashcards, visualizations, or custom study formats. It analyzes pickleball match footage to identify improvement areas and create training plans. AI Mode in Search uses Gemini 3 to generate visual layouts, interactive tools, and simulations on demand based on queries.
https://blog.google/products/gemini/gemini-3/ - Google introduces Gemini 3, its most intelligent AI model, enhancing reasoning and multimodal capabilities. You can now access Gemini 3 across Google products like the Gemini app, AI Studio, and Vertex AI. Expect Gemini 3 Deep Think mode for Ultra subscribers soon, with more models to follow. Gemini 3 is state-of-the-art in reasoning, built to grasp depth and nuance — whether it’s perceiving the subtle clues in a creative idea, or peeling apart the overlapping layers of a difficult problem. Gemini 3 is also much better at figuring out the context and intent behind your request, so you get what you need with less prompting. It’s amazing to think that in just two years, AI has evolved from simply reading text and images to reading the room. And starting today, we’re shipping Gemini at the scale of Google. That includes Gemini 3 in AI Mode in Search with more complex reasoning and new dynamic experiences. This is the first time we are shipping Gemini in Search on day one. Gemini 3 is also coming today to the Gemini app, to developers in AI Studio and Vertex AI, and in our new agentic development platform, Google Antigravity — more below. Like the generations before it, Gemini 3 is once again advancing the state of the art. In this new chapter, we’ll continue to push the frontiers of intelligence, agents, and personalization to make AI truly helpful for everyone. We hope you like Gemini 3, we'll keep improving it, and look forward to seeing what you build with it. Much more to come!
https://infotechkeeda.com/google-gemini-3-technical-analysis-guide/ - Gemini 3 Deep Think represents Google’s response to OpenAI’s o1 reasoning model, providing enhanced deliberation for extraordinarily complex problems requiring extended reasoning chains, creative problem-solving, and iterative refinement. Performance Benchmarks: Humanity’s Last Exam: 41.0% (compared to 37.5% for standard Gemini 3 Pro) GPQA Diamond: 93.8% (Graduate-Level Google-Proof Q&A) ARC-AGI-2: 45.1% (with code execution, ARC Prize Verified) Distinctive Characteristics: Extended Reasoning Time: Allocates additional computational resources for complex queries, implementing chain-of-thought processes visible to end users. Iterative Refinement: The model employs multi-pass reasoning, generating initial hypotheses, evaluating logical consistency, and refining conclusions. Strategic Planning: Optimized for problems requiring long-term planning, multiple constraint satisfaction, and creative solution generation. Availability: Gemini 3 Deep Think is currently undergoing comprehensive safety evaluations and will be released to Google AI Ultra subscribers in the coming weeks (as of November 2025).

Noah Fact Check Pro

The draft above was created using the information available at the time the story first emerged. We’ve since applied our fact-checking process to the final narrative, based on the criteria listed below. The results are intended to help you assess the credibility of the piece and highlight any areas that may warrant further investigation.

Freshness check

Score: 10

Notes: The narrative is current, with the earliest known publication date of December 4, 2025, matching the article's publication date. ([en.wikipedia.org](https://en.wikipedia.org/wiki/Gemini_%28language_model%29?utm_source=openai))

Quotes check

Score: 10

Notes: No direct quotes were identified in the provided text, indicating potential originality or exclusivity.

Source reliability

Score: 7

Notes: The narrative originates from WebProNews, a reputable outlet. However, the lack of direct quotes and reliance on secondary sources may affect reliability.

Plausability check

Score: 9

Notes: The claims about Gemini 3 Deep Think's capabilities align with information from Google's official blog and other reputable sources. ([blog.google](https://blog.google/products/gemini/gemini-3?utm_source=openai))

Overall assessment

Verdict (FAIL, OPEN, PASS): PASS

Confidence (LOW, MEDIUM, HIGH): HIGH

Summary: The narrative is current and plausible, with no significant issues identified. The lack of direct quotes and reliance on secondary sources slightly affect reliability, but overall, the information appears accurate and timely.

AI
Google
Deep Think