Google
•
Gemini 2.5 Pro
•
Released 
June 2025

Google
Gemini 2.5 Pro

Google's reasoning flagship with a 1M-token context window and text, image, audio, video and PDF input for complex analysis.

Modality:
Text
Image
Audio
Video
PDF
model ID
google/gemini-2.5-pro

Output Speed *

101.81
tok/s

Intelligence Index *

0
/ 100

Context Window *

1048576
tokens

Input price

13.5
Anytoken

Output price

90
Anytoken
Gemini 2.5 Pro: Long-Context Reasoning Across Text, Video and Code Gemini 2.5 Pro is Google's reasoning-first model in the Gemini 2.5 family, positioned above Gemini 2.5 Flash for harder tasks. It always thinks before answering, accepts text, images, audio, video and PDFs within a 1,048,576-token context window, and returns text up to 65,536 tokens. That combination suits codebase analysis, long-document and video understanding, and multi-step agent workflows where answer quality matters more than minimum latency. Access Gemini 2.5 Pro via AnyAPI.ai

Performance

Reasoning and Coding Results Backed by a 1M-Token Window

Gemini 2.5 Pro is built for tasks where the model must reason before answering. For the June 2025 release, Google reports 86.4% on GPQA Diamond and 82.2% on Aider Polyglot; these are provider-reported figures, not independent measurements. They point to strong science reasoning and multi-language code editing, and the 1M-token window lets one model read a large repository or document first. In production, reserve it for hard analysis and coding steps, and plan for thinking-token overhead compared with lighter models.

Technical Specifications

What the model supports

Gemini 2.5 Pro accepts text, image, audio, video and PDF input but outputs text only. Its 1,048,576-token context window is separate from the 65,536-token output cap, so large inputs fit while single responses stay bounded. Thinking is always on and cannot be disabled; developers tune it with a thinking budget. Thinking tokens are billed as output, so account for them in latency and cost planning. Knowledge cutoff is January 2025.
Verified Specifications — 
Gemini 2.5 Pro
*
Input modalities
Text
Image
Audio
Video
PDF
output modalities
Text
Context window
1048576
 tokens
Maximum output tokens
65536
Reasoning
Yes
Knowledge cutoff
June 2025
Pricing (standard)
13.5
 AnyTokens in
 / 
90
 AnyTokens out

Quickstart

Sample code for Gemini 2.5 Pro

import requests

url = "https://api.anyapi.ai/v1/chat/completions"

payload = {
    "stream": False,
    "tool_choice": "auto",
    "logprobs": False,
    "model": "gemini-2.5-pro",
    "messages": [
        {
            "content": [
                {
                    "type": "text",
                    "text": "Hello"
                },
                {
                    "image_url": {
                        "detail": "auto",
                        "url": "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg"
                    },
                    "type": "image_url"
                }
            ],
            "role": "user"
        }
    ]
}
headers = {
    "Authorization": "Bearer AnyAPI_API_KEY",
    "Content-Type": "application/json"
}

response = requests.post(url, json=payload, headers=headers)

print(response.json())
import requests url = "https://api.anyapi.ai/v1/chat/completions" payload = { "stream": False, "tool_choice": "auto", "logprobs": False, "model": "gemini-2.5-pro", "messages": [ { "content": [ { "type": "text", "text": "Hello" }, { "image_url": { "detail": "auto", "url": "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg" }, "type": "image_url" } ], "role": "user" } ] } headers = { "Authorization": "Bearer AnyAPI_API_KEY", "Content-Type": "application/json" } response = requests.post(url, json=payload, headers=headers) print(response.json())
View docs
Copy
Code is copied
const url = 'https://api.anyapi.ai/v1/chat/completions';
const options = {
  method: 'POST',
  headers: {Authorization: 'Bearer AnyAPI_API_KEY', 'Content-Type': 'application/json'},
  body: '{"stream":false,"tool_choice":"auto","logprobs":false,"model":"gemini-2.5-pro","messages":[{"content":[{"type":"text","text":"Hello"},{"image_url":{"detail":"auto","url":"https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg"},"type":"image_url"}],"role":"user"}]}'
};

try {
  const response = await fetch(url, options);
  const data = await response.json();
  console.log(data);
} catch (error) {
  console.error(error);
}
const url = 'https://api.anyapi.ai/v1/chat/completions'; const options = { method: 'POST', headers: {Authorization: 'Bearer AnyAPI_API_KEY', 'Content-Type': 'application/json'}, body: '{"stream":false,"tool_choice":"auto","logprobs":false,"model":"gemini-2.5-pro","messages":[{"content":[{"type":"text","text":"Hello"},{"image_url":{"detail":"auto","url":"https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg"},"type":"image_url"}],"role":"user"}]}' }; try { const response = await fetch(url, options); const data = await response.json(); console.log(data); } catch (error) { console.error(error); }
View docs
Copy
Code is copied
curl --request POST \
  --url https://api.anyapi.ai/v1/chat/completions \
  --header 'Authorization: Bearer AnyAPI_API_KEY' \
  --header 'Content-Type: application/json' \
  --data '{
  "stream": false,
  "tool_choice": "auto",
  "logprobs": false,
  "model": "gemini-2.5-pro",
  "messages": [
    {
      "content": [
        {
          "type": "text",
          "text": "Hello"
        },
        {
          "image_url": {
            "detail": "auto",
            "url": "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg"
          },
          "type": "image_url"
        }
      ],
      "role": "user"
    }
  ]
}'
curl --request POST \ --url https://api.anyapi.ai/v1/chat/completions \ --header 'Authorization: Bearer AnyAPI_API_KEY' \ --header 'Content-Type: application/json' \ --data '{ "stream": false, "tool_choice": "auto", "logprobs": false, "model": "gemini-2.5-pro", "messages": [ { "content": [ { "type": "text", "text": "Hello" }, { "image_url": { "detail": "auto", "url": "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg" }, "type": "image_url" } ], "role": "user" } ] }'
View docs
Copy
Code is copied
View docs
Code examples coming soon...

Limitations & Trade-offs

Where Gemini 2.5 Pro falls short

1
Mandatory thinking overhead. Gemini 2.5 Pro always reasons: the minimum thinking budget is 128 tokens and thinking cannot be switched off, unlike Gemini 2.5 Flash. Thinking tokens are billed as output, so simple requests such as classification or short chat replies carry avoidable latency and cost. For high-volume or latency-sensitive paths, use Gemini 2.5 Flash or Flash-Lite instead.
2
Text-only output. The model accepts images, audio, video and PDFs, but it returns text only, with a 65,536-token cap per response. It cannot generate images or audio, and very long documents or code files may need to be produced in chunks. For image generation or speech output, choose a dedicated model.
3
Costly long-context requests. Google prices prompts above 200K tokens at a higher tier, and large prompts add latency. Routinely filling the 1M-token window is expensive for chat-style or high-frequency traffic. For repeated questions over the same corpus, consider context caching or retrieval-based RAG rather than resending everything.
4
Dated knowledge and lifecycle. The knowledge cutoff is January 2025, so recent libraries, APIs and events need search grounding or supplied context. Gemini 2.5 Pro is also not the newest Gemini generation. Google documents model lifecycle on its deprecations page, so confirm availability and evaluate newer Gemini models before committing a new long-term build.

Best-Fit Workloads

Where this model earns its place

01
Repository-scale code analysis Google reports 82.2% on Aider Polyglot and 59.6% on SWE-bench Verified, and the 1M-token window can hold large codebases in a single request. This fits code review, cross-file refactoring plans and debugging in coding agents. These figures are provider-reported, and thinking latency makes it a poor fit for inline autocomplete. Use Flash for lightweight edits.
02
Long video, audio and document understanding Native input of video, audio, images and PDFs within one 1M-token context lets a single request cover lecture recordings, long reports or mixed-media case files. Pair it with a response schema to extract structured data. Output is text only, and very large multimodal prompts raise cost and latency, so cache repeated corpora.
03
Agent planning and orchestration The model supports function calling, code execution, Google Search grounding and URL context, with a tunable thinking budget for multi-step planning. It works well as the planner or verifier in an agent loop. Cheaper models can run routine tool calls. Because thinking cannot be disabled, running Pro on every step adds overhead.
04
Scientific and technical reasoning Google reports 86.4% on GPQA Diamond, a graduate-level science benchmark. This supports research assistants, technical Q&A and analysis of dense papers or specifications. Treat the score as provider-reported, validate on your own domain data, and use search grounding for topics after the January 2025 knowledge cutoff.
Pricing in anytokens via AnyAPI
Input
13.5
₳
Output
90
₳
Cache write
2.25
₳
Cache read
0.75
₳

Integration

Access Gemini 2.5 Pro via AnyAPI.ai

Access Gemini 2.5 Pro through AnyAPI.ai using a unified API built for multi-model AI applications. Integrate Gemini 2.5 Pro without maintaining a separate provider-specific connection, and keep the flexibility to test, switch, or combine models as your application requirements evolve.

01

One API integration

Access Gemini 2.5 Pro and other AI models through the same API workflow instead of maintaining separate integrations for every provider.

02

Easy model switching

Test Gemini 2.5 Pro against alternative models or switch models as your performance, capability, or cost requirements change without rebuilding your application around another provider API.

03

Flexible for production

Use Gemini 2.5 Pro from experimentation through production while keeping your AI stack flexible as workloads, traffic, and model requirements evolve.

04

Multi-model applications

Use Gemini 2.5 Pro for the workloads where it performs best and combine it with other models for tasks that require different capabilities, performance, or efficiency.

Frequently Asked Questions

Answers to common questions about integrating and using this AI model via AnyAPI.ai

Gemini 2.5 Pro supports a 1,048,576-token context window and up to 65,536 output tokens. The context window covers everything you send, including text, images, audio, video and PDFs, while the output limit caps the generated response. Google prices prompts above 200K tokens at a higher tier.

Yes, Gemini 2.5 Pro accepts text, image, audio, video and PDF input. Its output is text only. It cannot generate images or audio, so pair it with another model if your workflow needs those outputs.

No. Thinking is always on in Gemini 2.5 Pro. You can tune it with a thinking budget from 128 to 32,768 tokens or use dynamic thinking, but you cannot set it to zero. Gemini 2.5 Flash allows thinking to be disabled, which makes it better for simple, latency-sensitive requests.

Gemini 2.5 Pro is strong for complex coding by Google's own figures, which are 82.2% on Aider Polyglot and 59.6% on SWE-bench Verified. These are provider-reported. Its 1M-token context helps with whole-repository analysis. For high-volume or latency-sensitive edits, Gemini 2.5 Flash is usually the more practical choice.

Gemini 2.5 Pro sits in the premium reasoning tier of Google's lineup. Thinking tokens are billed as output, and prompts above 200K tokens use a higher pricing tier. Current AnyAPI.ai pricing appears in the AnyTokens pricing component on this page. For cost-sensitive volume workloads, Gemini 2.5 Flash is the lower-cost sibling.

* Benchmark data source: Artificial Analysis artificialanalysis.ai