Gemini 2.5 Pro

Name: Gemini 2.5 Pro
Brand: Google DeepMind
SKU: gemini-2-5-pro
Price: 1.5 USD
Availability: InStock

Tekst ja vestlusVanem versioonSaadaval

Google DeepMind pooltgemini-2-5-pro

Google's latest thinking model. Excels at reasoning, coding, math, and science with massive context window.

Proovi Hangi API võti Võrdle

Hind · 1M sisse/välja: 1,50 $ / 12,00 $
Kontekst: 1 000 000 tokenit
Maks. väljund: 65 536 tokenit
Sisend → väljund: Tekst → Tekst
Arendaja: Google DeepMind
Uuendatud: 23. september 2026

See on mudeli vanem versioon.

Playground

Proovi Gemini 2.5 Pro

Vestlus

1,50 $/1M sis

Saada sõnum. Vastus saabub täielikult, kui mudel on valmis (ilma voogesituseta).

Maks. vastuse pikkus (tokenid)

⌘/Ctrl + Enter

See käivitus

maksimaalselt 0,0123 $ · 1,23 krediiti reserveeritud

Arveldatakse tegelikult kasutatud tokenid; reserveeringu kasutamata osa tagastatakse.

Oled siin uus?

10 tasuta krediiti (0,10 $), kui registreeritud Google'iga

Kasutatav 24 tundi pärast registreerimist, kuni 5 käivitust päevas ja maksimaalselt 2 krediiti käivituse kohta. Muud sisselogimismeetodid algavad ilma krediitideta. Piisab 8 käivituseks sellest mudelist.

Teave: Gemini 2.5 Pro

LühidaltSeisuga 23. september 2026

Gemini 2.5 Pro on mudel, mille on välja töötanud Google DeepMind kategoorias Tekst ja vestlus. Railwail'is maksab Gemini 2.5 Pro 1,50 $ 1M sisendtokeni kohta ja 12,00 $ 1M väljundtokeni kohta. Konteksti aken sisaldab 1 000 000 märksõna ja üks vastus võib olla kuni 65 536 märksõna pikkune.

Taust

Teave Google DeepMind kohta

Asutatud 2010 · Mountain View, USA / London, UK

Google DeepMind is the merged AI research organisation formed in April 2023 by combining Google Brain (founded inside Google in 2011) with DeepMind (founded in London in 2010 by Demis Hassabis, Shane Legg and Mustafa Suleyman, acquired by Google in 2014). Demis Hassabis leads the unit as CEO. DeepMind authored seminal papers including 'Attention Is All You Need' (Google Brain, 2017), AlphaGo (2016), AlphaFold (2018-2021, awarded the 2024 Nobel Prize in Chemistry), AlphaZero, Chinchilla scaling laws and the Gemini Technical Report. The Gemini family launched in December 2023 (Ultra, Pro, Nano), followed by Gemini 1.5 Pro with 1M+ context and the Mixture-of-Experts architecture in early 2024, Gemini 2.0 Flash (December 2024) and Gemini 2.5 Pro (released March 2025, generally available May 2025) which introduced a built-in 'Deep Think' reasoning mode. Google DeepMind also ships Imagen, Veo, Lyria, NotebookLM and powers AI features across Google Search, Workspace and Android.

Külasta Google DeepMind

Arhitektuur

Sparse Mixture-of-Experts Transformer (natively multimodal, reasoning)

Gemini 2.5 Pro is Google DeepMind's flagship reasoning model released as preview in March 2025 and generally available in May 2025. It is a natively multimodal Sparse Mixture-of-Experts (MoE) Transformer that ingests text, image, audio and video tokens through a shared embedding space, building on the architecture introduced in Gemini 1.5 Pro. Pretraining used Google's TPU v5p infrastructure on a multi-trillion-token corpus mixing web text, code, books, scientific papers, image-text pairs, audio and YouTube video frames, with a knowledge cutoff of January 2025. Post-training combined supervised fine-tuning, RLHF and reinforcement learning against verifiable rewards for math and code, plus a new 'Deep Think' training stage that teaches the model to allocate test-time thinking budgets and emit long internal chains-of-thought before the final answer. Deep Think is exposed to developers as a toggle and was the headline capability at launch, pushing Gemini 2.5 Pro to the top of the Chatbot Arena leaderboard for the first time for Google. The model supports a 1,048,576-token context window with native handling of long PDFs, hour-long videos and full code repositories. Tool use, function calling, structured output and grounding with Google Search are first-class. Safety training included red-teaming under Google's Frontier Safety Framework.

Parameetrid: Undisclosed (sparse MoE, total parameters in the hundreds of billions, active per-token undisclosed)
Kontekst: 1 048 576 tokenit

Võimalused

1,048,576 token context window (~1M tokens) with native long-video and long-audio support
Built-in Deep Think reasoning mode with adjustable thinking budget
Natively multimodal: text, image, audio and video in a single pass
Top scores on GPQA Diamond, AIME 2025 and Humanity's Last Exam
Strong coding performance on SWE-bench Verified and LiveCodeBench
Function calling, JSON schema, parallel tool calls
Grounding with Google Search built into the API
Native execution of code via the Code Execution tool
Cross-lingual reasoning across 100+ languages
Available via Vertex AI, AI Studio, and the Gemini app
Best for: scientific reasoning, long-document and long-video analysis, coding, agentic workflows.

Treening ja litsents

Pretrained on a multi-trillion-token mixture of web text, code, books, scientific papers, licensed third-party text, audio waveforms, image-text pairs and video frames. Knowledge cutoff is January 2025. Post-training uses supervised fine-tuning, RLHF, RL against verifiable rewards and Deep Think reasoning training.

Litsents: Proprietary commercial license via Google AI Studio, Vertex AI and the Gemini app. Commercial use permitted under Google Cloud terms.

Turvalisuse testimine: Evaluated under Google DeepMind's Frontier Safety Framework v2 covering CBRN, cyber, persuasion and ML R&D risks, with internal red teams and external evaluators.

Teadaolevad piirangud

Deep Think adds significant latency and token cost
Long context recall quality degrades beyond ~500K tokens for some tasks
Vision occasionally misreads dense tables and handwriting
Knowledge cutoff January 2025
Region availability varies and is not yet in all EU regions for some features

Teadusartiklid

Hinnad

Hinnad USA dollarites. Kasutamine arvestatakse eelmakstud krediitidest.

Sisend	1,50 $ / 1M tokenit
Väljund	12,00 $ / 1M tokenit
Sisend (päringud üle 200 000 tokeni)	3,00 $ / 1M tokenit
Väljund (päringud üle 200 000 tokeni)	18,00 $ / 1M tokenit

Arvestatakse tokened, mida iga päring tegelikult kasutab.
1 krediit = 0,01 $

Kulukalkulaator

Hinna kalkulaator

Sisendtokened

/ päring

Väljundtokened

/ päring

Päringud

Kokku

0,75 $

75 krediiti

Päringute kohta

0,0075 $ · 0,75 krediiti

Iga päring ümardatakse 0,01 krediidini.

API

Kutsu Gemini 2.5 Pro oma Railwail API võtmega. Kasuta selles päringus seda mudeli ID-d:

gemini-2-5-proVaata API dokumentatsiooni Hangi API võti

curl https://railwail.com/api/v1/chat/completions \
  -H "Authorization: Bearer $RAILWAIL_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-2-5-pro",
    "messages": [
      {
        "role": "user",
        "content": "Explain what a vector database is in two sentences."
      }
    ],
    "max_tokens": 1024
  }'

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["RAILWAIL_API_KEY"],
    base_url="https://railwail.com/api/v1",
)

completion = client.chat.completions.create(
    model="gemini-2-5-pro",
    messages=[
        {
            "role": "user",
            "content": "Explain what a vector database is in two sentences.",
        },
    ],
    max_tokens=1024,
)

print(completion.choices[0].message.content)

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.RAILWAIL_API_KEY,
  baseURL: "https://railwail.com/api/v1",
});

const completion = await client.chat.completions.create({
  model: "gemini-2-5-pro",
  messages: [
    {
      role: "user",
      content: "Explain what a vector database is in two sentences."
    }
  ],
  max_tokens: 1024
});

console.log(completion.choices[0].message.content);

// npm install railwail
import railwail from "railwail";

const rw = railwail(process.env.RAILWAIL_API_KEY);

const res = await rw.chat("gemini-2-5-pro", [
  { role: "user", content: "Explain what a vector database is in two sentences." },
], { max_tokens: 1024 });

console.log(res.choices[0].message.content);

Määra oma võti kui RAILWAIL_API_KEYLoo API-võti

Spetsifikatsioonid

Mudeli ID: gemini-2-5-pro
Arendaja: Google DeepMind
Kategooria: Tekst ja vestlus
Sisend: Tekst
Väljund: Tekst
Konteksti aken: 1 000 000 tokenit
Maks. väljund: 65 536 tokenit
Arveldamine: Kasutuse järgi (tokenid või GPU-aeg)
Elutsükkel: Vanem versioon
Mudeli suurus: Undisclosed (sparse MoE, total parameters in the hundreds of billions, active per-token undisclosed)
Litsents: Proprietary commercial license via Google AI Studio, Vertex AI and the Gemini app. Commercial use permitted under Google Cloud terms.
Kataloogi kirje uuendatud: 23. september 2026

Sildid

reasoning
coding
multimodal

Näite-prompid

Näited Railwail kataloogist. Neid ei loodud otse sellel lehel.

Scientific Analysis
Explain CRISPR-Cas9 gene editing: how it works, current medical applications, and ethical considerations
Näita näite vastust
CRISPR-Cas9 works like molecular scissors guided by a GPS. A guide RNA is designed to match a specific DNA sequence; the Cas9 protein then cuts the DNA at that exact location, allowing scientists to delete, repair, or insert genetic material. Current medical applications include sickle cell disease treatment (Casgevy, approved 2023), where patients' stem cells are edited ex vivo to produce functional hemoglobin. Clinical trials are underway for certain cancers, hereditary blindness, and HIV resistance. Ethical considerations center on germline editing (changes passed to future generations), equitable access to expensive therapies, and the slippery slope toward 'designer babies'—most scientists support a moratorium on heritable human genome editing while governance frameworks catch up.
Long Context Analysis
I have a 200-page legal contract. Identify all clauses related to intellectual property ownership, data retention, and termination conditions.
Näita näite vastust
I can process your entire 200-page contract within my context window. Here's what I would systematically extract: 1) IP Ownership clauses: Look for sections titled 'Intellectual Property,' 'Work Product,' or 'Ownership of Deliverables'—these define who owns code, designs, and inventions created during the engagement. 2) Data Retention: Sections covering 'Data Handling,' 'Confidentiality,' or 'Information Security' typically specify how long each party retains data post-termination, deletion timelines, and compliance with regulations like GDPR. 3) Termination Conditions: Usually in a dedicated 'Termination' section, covering for-cause vs. for-convenience termination, notice periods, cure periods for breaches, and post-termination obligations. Upload the document and I'll provide exact clause references, page numbers, and a summary of each relevant provision with any potential conflicts between sections.

Kasutamise juhtumid

Milleks seda kasutatakse

Long-video and long-PDF analysis
Scientific research and literature review
Coding agents with Deep Think
Search-grounded chatbots
Multimodal data extraction
Enterprise document understanding

Seotud juhendid

Best Large Language Models

Kõik mudelid: Tekst ja vestlus

Korduma kippuvad küsimused

Mis on Gemini 2.5 Pro?

Gemini 2.5 Pro on mudel, mille töötas välja Google DeepMind kategoorias Tekst ja vestlus. Railwail'is saad seda kutsuda API võtme abil Railwail API kaudu.

Kui palju maksab Gemini 2.5 Pro Railwail'is?

Railwail'is maksab Gemini 2.5 Pro 1,50 $ 1M sisendtokeni kohta ja 12,00 $ 1M väljundtokeni kohta. Sulle arvestatakse see, mida iga päring tegelikult kasutab. Kasutamine makstakse eelnevalt ostetud krediitidest; 1 krediit võrdub 0,01 $.

Kui suur on Gemini 2.5 Pro konteksti aken?

Gemini 2.5 Pro konteksti aken sisaldab 1 000 000 märksõna. Üks vastus võib olla kuni 65 536 märksõna pikkune.

Kui kiire on Gemini 2.5 Pro?

Gemini 2.5 Pro kohta ei ole Railwail'is veel piisavalt mõõdetud käivitusi, et käivitusaega väita. See sõltub sisendist, seadistustest ja pakkuja koormusest.

Kas Gemini 2.5 Pro on parem kui Claude Fable 5.1?

See sõltub ülesandest. Gemini 2.5 Pro (Google DeepMind) ja Claude Fable 5.1 (Anthropic) on mõlemad mudelid kategoorias Tekst ja vestlus. Võrdlusleht näitab nende hindu ja spetsifikatsioone kõrvuti.

Võrdle Gemini 2.5 Pro ja Claude Fable 5.1

Kuidas kasutada Gemini 2.5 Pro API kaudu?

Loo Railwail API võti ja saada oma päring mudeli ID-ga gemini-2-5-pro. Koodinäited curl'i, Pythoni ja JavaScripti jaoks on selle lehe API jaotises.

Võrreldavad mudelid

Kõik selles kategoorias

Claude Fable 5.1Anthropic
Anthropic's model for the most demanding reasoning and long-horizon agentic work. 1M-token context window, up to 128K output tokens, adaptive thinking that is always on.
12,00 $/1M sis
700 % kallim ühiku kohta
Võrdle Gemini 2.5 Pro ja Claude Fable 5.1
Claude Opus 4.8Anthropic
The most capable model of Anthropic's Opus 4 series. State of the art on long-horizon agentic work, coding and knowledge tasks, with a 1M-token context window at standard pricing.
6,00 $/1M sis
300 % kallim ühiku kohta
Võrdle Gemini 2.5 Pro ja Claude Opus 4.8
Claude Opus 5.5Anthropic
Anthropic's current Opus model for long-running agentic coding and knowledge work. 1M-token context window, up to 128K output tokens, adaptive thinking that is always on.
4,80 $/1M sis
220 % kallim ühiku kohta
Võrdle Gemini 2.5 Pro ja Claude Opus 5.5

Kasuta Gemini 2.5 Pro API kaudu

Üks API võti kõigile Railwail mudelitele. Kasutamine arvestatakse eelmakstud krediitidest, 1 krediit = 0,01 $.

Hangi API võti Vaata kõiki mudeleid