Text model

Grok by xAI: a 2M token context and live web search

Q: Which Grok models are available in GPTunneL?

The current list is always visible in the «Every Grok model» block above — it is built straight from the platform catalogue.

Q: How much text can Grok handle?

Up to two million tokens on Grok 4.20 — thousands of pages in one conversation. The other models in the family have smaller windows; the exact figure is in the table above.

Q: Can Grok create images?

Yes, the current models in the family draw pictures from a text description right in the conversation — no separate model needed.

Q: Does Grok have access to X (Twitter) data?

There is no direct integration with the X platform in GPTunneL. The model gets fresh information from web search.

Q: How is Grok different from ChatGPT?

They are different tools. Grok has a much larger context and a freer manner of answering; the GPT family has more tiers for different budgets and a separate branch for code. The choice depends on the task, and both are available in GPTunneL.

Grok models by xAI in GPTunneL: context up to 2M tokens, web search and image generation. Pay per token, no subscription and no VPN.

Try Grok Available models

What the Grok family is good at

Traits xAI keeps from one generation to the next — you can count on them whichever model you pick.

The largest context

Up to two million tokens — more than any other family in the service. A whole library of documents fits into one conversation.

Fresh information

The model searches the web and leans on what it finds, so it can answer questions about things that happened recently.

Images in the same window

Grok draws pictures from a text description right in the conversation — no separate model needed.

A free-form manner

It answers more directly and with more life than most competitors, and dodges the topic with formal caveats less often.

Which model to pick

The three models in the family differ in context size and in the character of the answer.

	Grok 4.20Maximum context	Grok 4.3The default pick	Grok 4.5The newest
Best for	Very long piles of documents and correspondence	Everyday tasks, writing, analysis	Hard reasoning and work with images
Context	2M tokens	1M tokens	500K tokens
Reasoning	No	Yes	Yes
Image generation	No	Yes	Yes

Figures come from the platform catalogue. Exact per-model prices are on the Pricing

Every Grok model

Release dates follow xAI's official announcements.

July 2026

Grok 4.5

The current model in the family — stronger than its predecessors at reasoning and at working with images.

April 2026

Grok 4.3

A million tokens of context and reasoning the model switches on by itself when the task is hard.

February 2026

Grok 4.20 Beta

A record two million tokens of context — for tasks that need a great deal of text held at once.

November 2025

Grok 4.1

An update to the fourth generation: answers got noticeably faster, and a separate fast version followed.

July 2025

Grok 4

The generation that brought the family up to the level of the top models on the market.

Full release history

February 2025

Grok 3

The first generation with reasoning of its own and a clear jump in maths and code.

August 2024

Grok 2

The model learned to work with images and gained a smaller, cheaper version.

November 2023

Grok 1

xAI's first model — free-form in tone and connected to fresh data from the very start.

How billing works

There is no subscription: you top up one balance and spend it on any model on the platform.

Pay per token

You are charged for exactly what the request and the answer used. Not using it costs nothing: no limits, no monthly fee.

One balance for every model

Grok, ChatGPT, Claude, image and video generation — all out of the same wallet. No separate subscription per service.

Top up the way you prefer

International cards, Apple Pay, Google Pay or crypto. The minimum top-up is $5.

See pricing

How to get started with Grok

One account for every model on the platform.

Emailyou@example.comPassword••••••••

What was written this week about the new electricity tariffs? Give me a summary

Web search

Grok 4.5

Summary of the week

The main story of the week is the review of tariffs for private homes with electric heating.

Trade publications disagree: some expect a rise, others talk about the timeline slipping.

There are no official documents yet — every piece leans on comments from market participants.

The model switches right inside the input — you can change it mid-conversation.

The answer lands in the chat and the context is kept until the conversation ends.

Try Grok in GPTunneL

Signing up takes a minute, and $5 on the balance is enough to see whether the model fits your task.

Try Grok

Frequently asked questions

/01Do I need a VPN to use Grok through GPTunneL?

No. Requests go through GPTunneL's infrastructure, so Grok opens like any ordinary website — no VPN and no proxy.

/02Which Grok models are available in GPTunneL?

/03How much text can Grok handle?

/04Can Grok create images?

/05Does Grok have access to X (Twitter) data?

/06How is Grok different from ChatGPT?

Grok is a neural network from Elon Musk's company xAI and a direct rival to ChatGPT and Claude. It answers questions, writes and analyses text, works with code and draws images from a description — all in one conversation.

The family's defining feature is the size of its context: the larger models hold up to two million tokens, which is thousands of pages at once. Grok searches the web and leans on what it finds, so it answers questions about recent events and not only about what was in its training data. The tone is noticeably freer than with competitors: the model dodges the topic with formal caveats less often.

Every Grok model is available in GPTunneL — worldwide, without a VPN, with one balance and billing for the tokens you actually spend.