Enterprise AI with Command R in Turkey - The Unified AI API Platform

What Is Cohere Command R?

About the technology and the company behind the model

Cohere was founded in Toronto in 2019 by Aidan Gomez, Nick Frosst and Ivan Zhang. Gomez co-authored "Attention Is All You Need", the paper that introduced the Transformer architecture behind modern AI. Rather than chasing the consumer chatbot race, Cohere focused on enterprise AI from day one: its models are built to work safely with company data and can be deployed in the cloud, in private clouds or on-premises.

The current flagship, Command A, ships with a 256K-token context window and support for 23 languages including Turkish, and it ranks among the most efficient models in its class for enterprise RAG, search and agentic tasks. Command R+ 08-2024 remains the proven, balanced option for RAG and tool use, while the compact Command R7B 12-2024 handles high-volume classification, summarization and search workloads at low cost. What unites them is grounded generation: answers come with citations, so every claim can be traced back to its source document.

Onysoft AI Gateway makes Cohere API access from Turkey straightforward: you pay in Turkish lira from a TL balance converted at the current central bank (TCMB) rate, with up-to-date TL pricing, corporate e-invoicing and KVKK-compliant infrastructure. No foreign credit card or VPN is needed, and your existing OpenAI SDK code works after a single base_url change. The platform, operated by Izmir-based Onysoft Veri Merkezi A.S., puts 708+ models behind one OpenAI-compatible API, with a Playground for testing and 24/7 support in Turkish.

Cohere's roadmap centers on agentic systems that embed models deep inside enterprise workflows. For teams investing in auditable, citation-backed AI, the Command family is a solid foundation today and a safe bet for what comes next.

business

Provider

Cohere

calendar_month

Founded

2019

location_on

Headquarters

Toronto, Canada

Command R Performance Metrics

Results from industry-standard benchmark tests

MMLU Multi-Domain Knowledge 80.2

HumanEval Code Generation 75.0

RAG Score RAG Performance 94.0

Tool Use Tool Use 91.5

* Benchmark results are based on data published by the respective providers and may be updated.

Why Choose Command R?

Built for Enterprise RAG and Search

search

RAG Optimization

The Command family is trained specifically for retrieval-augmented generation, producing verifiable, consistent answers grounded in your own documents.

source

Built-In Citations

Answers arrive with inline citations mapping each claim to its source document — ideal for work that demands auditability and compliance.

build

Tool Use & Agentic

Advanced function calling and multi-step tool use. Build agents that query databases, call APIs and drive complete workflows.

language

23 Languages Including Turkish

Command A is trained on 23 languages including Turkish. Run multilingual customer support, document processing and global operations on one model.

data_object

Up to 256K Context

Command A offers 256K tokens of context; Command R+ and R7B offer 128K. Process long contracts, reports and document sets in a single request.

security

Enterprise Security & Compliance

Cohere is engineered enterprise-first for data security; through Onysoft you add KVKK-compliant infrastructure, TL billing and corporate e-invoicing.

Command R Models and Pricing

Command R models available in TL through Onysoft AI Gateway

Recommended

Command A

Flagship

Cohere's flagship. 256K context, 23 languages including Turkish — for the hardest RAG, search and agentic tasks.

cohere/command-a play_circle Try

arrow_downward Input: ₺0,18 / 1K tokens

arrow_upward Output: ₺0,71 / 1K tokens

Command R+ (08-2024)

Popular

Proven RAG and tool-use performance. The strong enterprise choice for balancing quality and cost.

cohere/command-r-plus-08-2024 play_circle Try

arrow_downward Input: ₺0,18 / 1K tokens

arrow_upward Output: ₺0,71 / 1K tokens

Command R7B (12-2024)

Budget

Compact and fast. The most economical pick for high-volume classification, summarization and search workloads.

cohere/command-r7b-12-2024 play_circle Try

arrow_downward Input: ₺0,00 / 1K tokens

arrow_upward Output: ₺0,01 / 1K tokens

view_list View All Models and Pricing

Command R Use Cases

What businesses in Turkey can build with Command R

manage_search

Enterprise Knowledge Management

Make internal documents, wikis and knowledge bases searchable with citation-backed answers, so employees reach the right information in seconds.

support_agent

Customer Support Automation

Build support assistants that answer from your product docs with citations. Handle high ticket volumes at low cost with R7B.

gavel

Legal & Compliance

Analyze contracts, regulations and internal policies within a 256K context, and document the exact clause behind every answer with citations.

analytics

Financial Analysis & Reporting

Summarize long financial reports, balance-sheet notes and market documents while keeping every numeric claim traceable to its source.

integration_instructions

Agentic Workflows

Use tool support to build agents that call your CRM, ERP and internal APIs — automations that open tickets, update records and compile reports.

description

Multilingual Document Processing

Translate, summarize and classify in 23 languages including Turkish. Manage the document flow of multinational operations with a single model.

Which Sectors Does Command R Serve?

Sectors using Command R through Onysoft AI Gateway

Enterprise Legal Finance Healthcare Insurance Public Sector Consulting HR

How to Get Started with Command R?

Start using Command R models in your project in 3 steps

1

Create an Account

Sign up to Onysoft AI Gateway for free. Sign in instantly with your email address or your GitHub/Google account. Your API key is generated automatically.

2

Add Balance

Add as much balance as you like in Turkish Lira. You can pay by bank transfer/EFT, credit card or corporate invoice. Transparent TL pricing with the current TCMB exchange rate.

3

Start Using the API

Thanks to our OpenAI-compatible API endpoint, access Command R models without changing your existing code. Just change the base_url and api_key.

from openai import OpenAI

client = OpenAI(
    api_key="sk-ony-...",
    base_url="https://api.onysoft.com/v1"
)

response = client.chat.completions.create(
    model="cohere/command-a",
    messages=[{"role": "user", "content": "Merhaba!"}]
)
print(response.choices[0].message.content)

import OpenAI from 'openai';

const client = new OpenAI({
    apiKey: 'sk-ony-...',
    baseURL: 'https://api.onysoft.com/v1'
});

const response = await client.chat.completions.create({
    model: 'cohere/command-a',
    messages: [{role: 'user', content: 'Merhaba!'}]
});
console.log(response.choices[0].message.content);

curl https://api.onysoft.com/v1/chat/completions \
  -H "Authorization: Bearer sk-ony-..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "cohere/command-a",
    "messages": [{"role": "user", "content": "Merhaba!"}]
  }'

$ch = curl_init('https://api.onysoft.com/v1/chat/completions');
curl_setopt_array($ch, [
    CURLOPT_HTTPHEADER => [
        'Authorization: Bearer sk-ony-...',
        'Content-Type: application/json'
    ],
    CURLOPT_POSTFIELDS => json_encode([
        'model' => 'cohere/command-a',
        'messages' => [['role' => 'user', 'content' => 'Merhaba!']]
    ]),
    CURLOPT_RETURNTRANSFER => true,
    CURLOPT_POST => true
]);
$response = json_decode(curl_exec($ch));
echo $response->choices[0]->message->content;

Why Use Command R with Onysoft?

The advantages of choosing Onysoft AI Gateway over the Cohere API directly

currency_lira

Pay in Turkish Lira

No foreign credit card needed. Transparent pricing in TL with the current TCMB exchange rate. Bank transfer, credit card or corporate invoice.

receipt_long

Invoicing in Turkey

As a company based in Turkey, we issue e-invoices and e-archive invoices. Your accounting processes run smoothly.

hub

One API, All Models

Access 370+ models including Command R just by changing the base_url. OpenAI SDK compatible.

support

24/7 Support in Turkish

Our Turkish-speaking technical support team is always with you. Integration help and issue resolution.

shield

KVKK Compliant

Data processing policies aligned with Turkish regulations. Enterprise data security standards.

trending_down

Cost Optimization

Switch easily between different models as your needs change. Optimize your budget.

Frequently Asked Questions About Command R

What people wonder about Command R and Onysoft AI Gateway

How can I access the Cohere API from Turkey? expand_more

Sign up on Onysoft AI Gateway, top up a TL balance and you get instant access. No foreign credit card, VPN or overseas billing address is needed — just point your requests at https://api.onysoft.com/v1.

How does Cohere API pricing in Turkish lira work? expand_more

It is pure pay-as-you-go — no subscription, no commitment. Token-based charges are converted to TL at the current TCMB exchange rate and deducted from your balance. Up-to-date TL pricing for each model is shown in the pricing table on this page.

What about KVKK compliance and data privacy? expand_more

Onysoft AI Gateway runs KVKK-compliant infrastructure; your account and billing data are processed under Turkish data protection law. Cohere itself is an enterprise-first provider with a strong data security posture. For corporate data processing details, our 24/7 support team can walk you through the specifics.

Which Cohere model should I pick, and when? expand_more

Choose Command A for top quality, long context and agentic tasks. Command R+ 08-2024 is the proven pick for RAG projects that balance quality and cost. For high-volume classification, summarization and search, the compact Command R7B 12-2024 is the most economical — and you can compare all three in the Playground.

What is Command A and what makes it stand out? expand_more

Command A is Cohere's current flagship: a 256K-token context window, support for 23 languages including Turkish, and strong RAG and tool-use capabilities. Its efficient architecture delivers high throughput for its class, and it is designed specifically for enterprise agentic workloads.

What are the context window limits? expand_more

Command A offers a 256,000-token context window, while Command R+ 08-2024 and Command R7B 12-2024 both support 128,000 tokens. In practice that means long contracts, reports and document collections can be processed in a single request.

Can I use Cohere with my existing OpenAI SDK code? expand_more

Yes. Set base_url to https://api.onysoft.com/v1, plug in your Onysoft API key and pass a model id like cohere/command-a. The rest of your code — including streaming — works unchanged.

Do you issue corporate invoices? expand_more

Yes. Every balance top-up is documented with a corporate e-invoice (e-fatura) issued to your company. Your Cohere API spend flows straight into your accounting and expense processes, fully compliant with Turkish regulations.

What is RAG, and why is Cohere strong at it? expand_more

RAG (Retrieval-Augmented Generation) means the model grounds its answers in information retrieved from your own documents. Command models are trained specifically for RAG: they return citations showing which claim came from which document. That reduces hallucination risk — a critical advantage for organizations that need auditability.

Do Cohere models support Turkish? expand_more

Yes. Command A is trained on 23 languages including Turkish, and the other Command models also include Turkish in their training data. They handle Turkish document search, summarization and customer-support scenarios comfortably.

scatter_plot Cohere Command R