Text model

The Claude family — general-purpose LLMs by Anthropic

Q: Which Claude models are available in GPTunneL?

The current list is always visible in the «Every Claude model» block above — it is built straight from the platform catalogue. Usually that means the top models of the generation (Opus, Fable), the general-purpose Sonnet and the fast Haiku.

Q: Which Claude model should I use for work?

For most tasks — Sonnet: it is fast and much cheaper than the larger models. If the task is hard (a big refactor, a long agent workflow, a pile of documents to read), go for Opus or Fable. For a stream of small repetitive requests Haiku is the cheapest option.

Q: How large is Claude's context?

The current top models hold 1M tokens — roughly 500 thousand words: an entire codebase or a folder of documents in one conversation. The fast Haiku has a smaller window, 200 thousand tokens. The exact figure for each model is in the list above.

Q: Which languages does Claude handle?

The family works comfortably in dozens of languages — it understands the request and answers in the same language, including long texts and documents.

Q: Can I send Claude images and PDFs?

Yes, the models accept images and files: diagrams, screenshots, tables and scanned documents are read together with the text of your request.

All Claude models in GPTunneL: Opus, Sonnet, Haiku and Fable. Context up to 1M tokens, pay per token, no subscription and no VPN.

Try Claude Available models

What the Claude family is good at

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

Context up to 1M tokens

The larger models hold an entire codebase, a contract with all its annexes or a folder of research papers in a single conversation — no need to cut it into pieces.

Code and agent workflows

Claude calls tools and keeps long multi-step tasks on track, which is why the family is a default pick for development and automation.

Careful wording

The models push back on a wrong premise more often and are readier to say «I don't know» — that matters when the answer goes to a client or into a document.

Four tiers for the job and the budget

Haiku for a stream of small requests, Sonnet for everyday work, Opus and Fable for the hardest tasks. Switching takes one click.

Which model to pick

Claude comes in four tiers — for different jobs and budgets.

	HaikuVolume and speed	SonnetThe default pick	OpusHard tasks	FableMaximum capability
Best for	Short answers, classification, moderation, translation	Writing, analysis, everyday development	Large projects, code review, agents	Long autonomous workflows
Context	200K tokens	1M tokens	1M tokens	1M tokens
Answer length	up to 64K tokens	up to 128K tokens	up to 128K tokens	up to 128K tokens
Thinking	Extended, on request	Adaptive	Adaptive	Adaptive, always on
Speed	Highest	High	Medium	Below average
Price	Lowest	Moderate	Above average	Highest

Figures reflect the current model at each tier. Exact per-model prices are on the Pricing

Every Claude model

Release dates follow Anthropic's official announcements.

July 2026

Claude Opus 5

The current flagship Claude — complex development, long agent workflows and enterprise tasks.

June 2026

Claude Fable 5 Claude Sonnet 5

Fable arrived as a new top tier for long autonomous work. Sonnet 5 caught up with Opus on agent tasks while staying much cheaper.

May 2026

Claude Opus 4.8

More accurate on code and steadier on tasks the model works through dozens of steps in a row.

April 2026

Claude Opus 4.7

A new tokenizer and a clear step up in reasoning quality on long context.

February 2026

Claude Sonnet 4.6Claude Opus 4.6

Context grew to 1M tokens and adaptive thinking arrived — the model decides for itself how long to think about a request.

Full release history

Autumn 2025

Claude Haiku 4.5Claude Sonnet 4.5Claude Opus 4.5

Sonnet 4.5, then Haiku 4.5 and Opus 4.5. For the first time the smallest model in the family came close to recent flagships.

May 2025

Claude Opus 4Claude Sonnet 4Claude Opus 4.1

The generation built for long tool-using tasks — the model learned to hold on to a goal through an entire multi-step scenario.

February 2025

Claude 3.7 Sonnet

Extended reasoning — on a hard task the model can think for longer instead of answering straight away.

2024

Claude 3.5 SonnetClaude 3.5 Haiku

The middle tier matched the previous flagship and became noticeably stronger at code. From then on Sonnet is the workhorse of the family.

March 2024

Claude 3 OpusClaude 3 SonnetClaude 3 Haiku

The three-tier split the family still uses appeared here — fast, general-purpose and most capable.

July 2023

Claude 2Claude 2.1

A 100,000-token context when competitors offered 8–32 thousand. This is the generation that made Claude the model for working with documents.

March 2023

Claude 1

Anthropic's first public model. From the start the bet was on long context and predictable behaviour rather than on the widest bag of tricks.

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

Claude, ChatGPT, Gemini, 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 Claude

One account for every model on the platform.

Emailyou@example.comPassword••••••••

Read the attached contract and list the risks for the contractor

Contract.pdf

Claude Sonnet

Contract review

The main risk is clause 7.2: the contractor is liable for delay with no cap on the amount.

Acceptance has no deadline, so the client can hold off signing the handover act indefinitely.

Only the client may terminate unilaterally — worth asking for a symmetrical right.

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 Claude in GPTunneL

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

Try Claude

Frequently asked questions

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

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

/02Which Claude models are available in GPTunneL?

/03Which Claude model should I use for work?

/04How large is Claude's context?

/05Which languages does Claude handle?

/06Can I send Claude images and PDFs?

Claude is a neural network built by the American company Anthropic and one of the main rivals to ChatGPT and Gemini. It works with text: answers questions, writes and edits material, reads documents, programs and reviews other people's code.

Claude is chosen where the answer has to be precise rather than impressive: contracts and financial reports, technical documentation, development and code review, analytics, customer support. The model holds up to a million tokens in context — about 500 thousand words, which means an entire codebase or a folder of documents in one conversation, with nothing cut into pieces. Claude understands images and PDFs: attach a scan, a diagram or a table and ask about it in plain text.

The family has several models for different jobs and budgets: the fast and cheap Haiku, the general-purpose Sonnet, the powerful Opus and Fable for the hardest work. All of them are available in GPTunneL — worldwide, without a VPN, with one balance and billing for the tokens you actually spend.