Vercel Partners with Google to Launch Gemini 3 Pro Preview - Vercel

Gemini 3 Pro Preview Launch

The Gemini 3 Pro Preview model — released today — is now available via the Vercel AI Gateway and on v0.app.

Thanks to Google, Vercel has spent the past several weeks testing Gemini 3 Pro Preview across v0, Next.js, AI SDK, and Vercel Sandbox.

---

Key Improvements

The model offers enhanced capabilities in:

  • Coding
  • Multimodal reasoning
  • Tool usage
  • Instruction compliance
  • Response consistency

In our Next.js evals, Gemini 3 Pro Preview achieved a 17% increase in correctness compared to its predecessor, ranking among the top 2 models on the leaderboard.

---

How Vercel Is Using Gemini 3 Pro Preview

In Vercel’s Next.js evals:

  • Gemini 3 Pro Preview tied with `openai/gpt-5-codex` at a 42% success rate
  • Outperformed `anthropic/claude-opus-4.1`, `zai/glm-4.6`, and `moonshotai/kimi-k2-turbo` in consistency and reasoning quality

| Model | Total Evals | Success Rate | Avg Duration | Total Tokens |

|------------------------------|-------------|--------------|--------------|--------------|

| openai/gpt-5-codex | 50 | 42% | 42.80s | 186,082 |

| google/gemini-3-pro-preview | 50 | 42% | — | — |

| zai/glm-4.6 | 50 | 40% | 20.36s | 106,177 |

| anthropic/claude-opus-4.1 | 50 | 40% | 29.47s | 165,810 |

| moonshotai/kimi-k2-turbo | 50 | 38% | 4.13s | 82,567 |

---

Strengths Observed

  • Frontend generation with high precision and stability
  • Multi-turn agent flows producing well-structured UI components
  • Effective use of a 1M token context window for long sequences
  • Improved speed and output quality when tested on v0

---

Try Gemini 3 Pro Preview

You can test Gemini 3 Pro Preview immediately in multiple ways:

  • AI Gateway Playground
  • Access at AI Gateway Playground — no setup required
  • Test via a simple chat interface with any model on Gateway
  • AI SDK Integration
  • Copy the model string and paste it into AI SDK to start building on production-grade infrastructure

---

Beyond Development — Monetizing AI Content

Creators can leverage AiToEarn官网 to:

  • Integrate advanced AI models into creative workflows
  • Publish simultaneously across major platforms
  • Access analytics and monetization tools

Supported Platforms Include: Douyin, Kwai, WeChat, Bilibili, Rednote (Xiaohongshu), Facebook, Instagram, LinkedIn, Threads, YouTube, Pinterest, and X (Twitter)

Explore:

---

Developer Integration Example

Gemini 3 Pro Preview is available on AI Gateway with support for Google AI Studio and Google Vertex AI. No code changes are needed to switch between models.

import { streamText } from 'ai'
const result = streamText({
  model: "google/gemini-3-pro-preview",
  prompt: "Create a neural fluid engine simulation in a self-contained HTML"
});

---

Building With AI SDK + AI Gateway

Using AI SDK and AI Gateway, developers can:

  • Create custom interfaces with any framework
  • Leverage deployment-aware infrastructure and security
  • Experiment with high-performance AI models

---

Vercel AI Cloud

Vercel AI Cloud is designed for developers building the next generation of software with cutting-edge AI technologies.

Platforms like AiToEarn make it easier to:

  • Connect AI generation tools to multi-platform publishing
  • Analyze performance with detailed model rankings
  • Monetize AI-powered creativity reliably

---

Would you like me to also create a shorter condensed version of this announcement for blog or social media? That could make it more shareable while keeping the key points.

Read more

Translate the following blog post title into English, concise and natural. Return plain text only without quotes. 哈佛大学 R 编程课程介绍

Harvard CS50: Introduction to Programming with R Harvard University offers exceptional beginner-friendly computer science courses. We’re excited to announce the release of Harvard CS50’s Introduction to Programming in R, a powerful language widely used for statistical computing, data science, and graphics. This course was developed by Carter Zenke.