Model Library
Browse and deploy state-of-the-art AI models through the DevUp Gateway.
Browse and deploy state-of-the-art AI models through the DevUp Gateway.
Gemini 2.5 Pro is Google's frontier-level multimodal model, engineered for highly complex reasoning, advanced coding, and massive-scale data analysis. Featuring an unprecedented 2M context window and hybrid thinking capabilities, it excels at solving the most demanding enterprise and scientific tasks.

Model Page: Gemini 2.5 Pro
Resources and Technical Documentation:
Authors: Google DeepMind
Gemini 2.5 Pro is Google's most capable and intelligent model to date. Designed as a frontier-class multimodal system, it is built to handle highly complex, multi-step reasoning, intricate codebase architectures, and deep scientific data analysis.
It utilizes an advanced Hybrid Reasoning Architecture that allows the model to allocate a massive computational budget to "think" before it speaks. Combined with a groundbreaking 2-million token context window, Gemini 2.5 Pro can ingest entire code repositories, hours of video, or vast libraries of documents in a single prompt without losing retrieval accuracy (Needle-in-a-Haystack accuracy approaches 100%).
The model operates on a Massive Sparse Mixture-of-Experts (MoE) architecture. Unlike dense models, it dynamically routes tokens to highly specialized expert neural networks during inference. This allows it to possess a colossal parameter count (enhancing deep knowledge and reasoning) while maintaining manageable compute costs. It was trained extensively on Google's next-generation TPUv5p clusters.
| Property | Limit |
|---|---|
| Maximum Input Tokens | 2,097,152 (2M context window) |
| Maximum Output Tokens | 8,192 (Standard) |
With its 2M token context window, Gemini 2.5 Pro pushes the boundaries of native multimodal ingestion.
Gemini 2.5 Pro can be seamlessly integrated using standard OpenAI-compatible SDKs. Specify the model string to leverage its advanced reasoning capabilities.
import OpenAI from "openai";
const openai = new OpenAI({
baseURL: "YOUR_BASE_URL", // Your designated API endpoint
apiKey: "YOUR_API_KEY",
});
async function runAdvancedReasoning() {
const completion = await openai.chat.completions.create({
model: "google/gemini-2.5-pro",
messages: [
{ role: "system", content: "You are a Principal AI Researcher. Output strict JSON format." },
{ role: "user", content: "Read these 50 academic PDFs and synthesize a novel mathematical theorem." }
],
temperature: 0.2,
});
console.log("Response:", completion.choices[0].message.content);
}
runAdvancedReasoning();