Text model

Cohere Command R — built for search over your documents

Q: Which Cohere models are available in GPTunneL?

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

Q: What is RAG and what does Cohere have to do with it?

RAG is when the model answers from your documents rather than from memory: the relevant passage is found in a base first, then put into the request. The Command R family was designed for exactly that, so it sticks to the text more carefully and shows its sources.

Q: How is Command R+ different from Command R?

Same family: R+ is the larger model, better on hard questions and multi-step scenarios but noticeably more expensive. R is the one to take for a stream of repetitive queries against a base.

Q: Which languages does Cohere handle?

The models are multilingual — they understand the request and answer in the same language.

Q: Should I use Cohere instead of ChatGPT?

It depends on the task: for open conversation, code and creative writing the general-purpose models are stronger. Cohere makes sense where you need a precise, citable answer from your own document base. Both families are available in GPTunneL.

Command R models by Cohere in GPTunneL: built for RAG and answers from your own document base, with citations. Pay per token, no subscription.

Try Cohere Available models

What the Command R family is good at

The traits that get this family picked for enterprise work.

Built for document work

The key difference from general-purpose models: this family was designed from the start for answering from your own knowledge base rather than for open-ended conversation.

Citations in the answer

The model shows which document each part of the answer came from. In enterprise settings that matters more than elegant phrasing — the answer can be verified.

Works with tools

Command R can call external tools and run multi-step scenarios, which is what internal assistants are built on.

128 thousand token context

Around two hundred pages of text per request — a set of policies or a large report fits in full.

Which model to pick

Two models in the family — they differ in size and price.

	Command RThroughput and volume	Command R+Hard tasks
Best for	Search over a base, fact extraction, simple answers	Hard questions over documents, multi-step scenarios
Context	128K tokens	128K tokens
Speed	Fast	Medium
Price per token	Low	High

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

Every Cohere model

Release dates follow Cohere's official announcements.

April 2024

Cohere Command R+

The larger model of the family: the same focus on document work, but noticeably better on hard questions and multi-step scenarios with tool calls.

March 2024

Cohere Command R

The model where Cohere committed to the enterprise scenario: answers from your own document base with citations, rather than open-ended conversation.

2023

Command

The first generation of the family, the one the company entered the business LLM market with.

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

Cohere, 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 start with Cohere

One account for every model on the platform.

Emailyou@example.comPassword••••••••

From these policies, tell me who approves a purchase below and above the new threshold, and cite the clauses

policy-2026.pdf

Cohere Command R+

Who approves a purchase

Below the new threshold the head of the unit approves it alone — clause 4.2 of the policy.

Above the threshold finance joins in, and beyond the annual limit a committee does — clause 4.5.

The old order still applies to requests registered before the change took effect — transitional clause 9.1.

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

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

Try Cohere in GPTunneL

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

Try Cohere

Frequently asked questions

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

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

/02Which Cohere models are available in GPTunneL?

/03What is RAG and what does Cohere have to do with it?

/04How is Command R+ different from Command R?

/05Which languages does Cohere handle?

/06Should I use Cohere instead of ChatGPT?

Cohere is a Canadian company that built its language models for business rather than for a mass audience from the outset. Its Command R family is tuned for one specific scenario: answering not from memory but from your own document base — finding the relevant passage, building the answer on it, and showing which file it came from.

That is why the family ends up in internal assistants and reference systems, where being able to verify an answer against its source matters more than elegant phrasing. The models can call external tools and run multi-step scenarios, and a 128 thousand token context covers roughly two hundred pages of text per request.

Cohere models are available in GPTunneL — worldwide, without a VPN, with one balance and billing for the tokens you actually spend.