Dhenara is app similar to ChatGPT which enables you to interact with all leading AI models in the world.
When you ask a question in ChatGPT, behind the scenes, it is responded by an LLM ( Large Language Model) which is like a 'super-brain'. In our platform, we support LLMs from multiple companies in the world, including the one those powers ChatGPT. You can choose and change the model you want for every single chat to get the best results.

Everybody can use Dhenara, no technical knowledge is required. All you need is an active subscription.

The plans comes with a fixed monthly usage limit ( $12 for Smart and $16 for Smart-Team). There are no daily or hourly limits, use whenever you want to use. Cost calculation is explained in the next question.

For every chat, cost is calculated based on 3 factors, the mode, number of words send to LLM,number of words generated by LLM.

Model Selection

Some models are expensive and some comes cheap. Below is a classification based on how expensive they are.

Super Expensive Models

• o1-preview
• dall-e-3 (Image Model)

Reasonably Expensive Models

• o1-mini
• gpt-4o
• claude 3.5 sonnet
• gemini-1.5-pro
• dall-e-2 (Image Model)

Inexpensive Models

• gpt-4o-mini
• claude 3.5 haiku
• gemini-1.5-flash

Number of words send to LLM

The next factor is the number of words in your chat or the content of attached files. More words you send to the LLM, more cost occurs. Remember, when you ask a follow up question at the end of an existing chat, all previous questions and answers of that chat will also be send to the LLM to give it the context of your new follow up question, and all of its words are counted towards cost.
The trick here is to stop lengthy follow up chats and start a new chat with fresh content whenever possible.

Number of words generated by LLM

You don't have direct control over this factor but you can always instruct the LLM with related instructions. For example, at the end of the chat you can add a sentence like Just give answer or Make output short . In general output words are 5x expensive than input words.

Exhausted reading about cost calculation ? Don't worry too much about it, that's the reason we show you the cost live after every chat. Just keep an eye on it. If you run out , you can always add Top ups which comes with infinite validity.

Yes, you can cancel subscription at any time. Note that no refund will be issued, but you will not be charged from the next subscription date onwards. After cancelling subscription, you can still use Dhenara till its last valid day.

Example: You bought a plan on 5th Dec. Subscription are auto renewed on 5th of every month. If you cancel the subscription on 10th Dec, you can continue to use the platform till 4th of Jan. On 5th Jan, no money will be taken from your card and you will loose access to the platform.

Go to My Profile pages, go to Subscriptions page. For Team plan, this page available only to workspace owners under Manage Workspace pages.

In the Subscriptions page, under Manage Subscriptions heading, click on the open_in_new . Here you can update your payment methods or cancel subscription.

Sorry that we won't be able to issue a refund once a payment is made. But you can always cancel the subscription whenever you want ( see above question to know details ) , and always resume if you change you mind later.

We use Stripe , which is the world's leading payment API platform and is absolutely safe. We do not store or access your payment method data such as card details, in our platform. If you choose to store it in the checkout, stripe will store it in their platform for any future payments.

No,
The only limit is your spend limit for the plan ( Eg: $10 for Smart Plan), and it applies for the entire month. Other than that, you are free to use any models at your own convenience.
We understand that it can be challenging to plan and organize usage. Perhaps one day you want to use it intensively for work, and the next day you may not want to use it at all. That's why we don't enforce hourly or daily limits. Also, if you run our of credits in any month, you can do a top-up and the top-up will never expire, you can use then as long as you have a valid subscription.

It's hard to point to a single model that fits all use cases. Your choice of models depends on what you are asking and how much you want to spend on that question.

Below is the list of models that we use and suggest for each use case. Note that cost is not the only factor that decides the quality of output. For example, for most of the programming questions, Claude 3.5 Sonnet or o1-mini models would even give better results than the expensive o1-preview model.

In most scenarios, we would recommend starting with a model, and then regenerating the same question with a different model and selecting the best answer that you decide. That is our primary feature: you can regenerate any node with any model just by clicking the autorenew button.
After regenerating, you can navigate through different choices using the <   > buttons shown at the end of the answer.

Programming	Claude 3.5 Sonnet o1-mini GPT-4o Gemini 1.5 Pro	Reasonable cost
General Purpose	Claude 3.5 Haiku GPT-4o-mini Gemini 1.5 Flash	Cheap, great for document analysis, simple programming questions, summarization, or anything involving very long input.
Reasoning / Extra Intelligence	o1-preview	Highly expensive

Right after every chat, you will be shown the latest usage data with remaining balance. You can also see detailed reports under My Usage . Worksapce owners can see workspace wide usage reports in Workspace Usage page.

No, in any of the plans, your data will not be used for model training. See next question for more details about your data with LLM providers.

Yes. All model providers do save the chats at their servers for a period raging from 30-60 days. This is purely for safety and safety related monitoring purposes and protect from abuse of technology. It will NOT be used for any other commercial purposes like training models or knowing user behaviour.

Absolutely No. We don't use or sell your data.

Currently we support

• Open-AI Chat Models
• gpt-4
• gpt-4o
• gpt-4o-mini
• gpt-4-turbo
• gpt-3.5-turbo
• Google Gemini Chat Models
• gemini-1.5-flash
• gemini-1.5-pro
• gemini-1.0-pro
• Anthropic Models
• claude-3.5-sonnet
• claude-3-sonnet
• claude-3-opus
• claude-3-haiku

To accessing Open-AI models via APIs, you have 2 options

• Through direct API call to models hosted in Open-AI Infra
• Through Microsoft API calls to the models hosted in Azure

Currently we support only Open-AI API keys, but if your organization have specific need to use only Microsoft hosted models, talk to us and we will enable it for you.

To create an OpenAI API key, visit OpenAI platform

Short Answer: Buy a plan from us, add API keys of the LLM providers into to your workspace and start chat/interactions.We are not providers of LLMs, we are only a platform where you can add your own API keys (securely) to your workspace and interact with them easily.

For Google models API access, you need to add it in aistudio.google.com . Note that you need to have GCP (Google-Clould-Platform) project with appropriate billing enabled for this. Read Gemini Documentation for more details.

To create an Anthropic API key, visit Anthropic developer platform

When you purchase a plan from us, you will get access to your workspace. Nothing else is included in the plan. For the LLM usage, you will directly pay to their providers.

Open AI

As per OpenAi documentation on 5th July 2024, they state,

At OpenAI, protecting user data is fundamental to our mission. We do not train our models on inputs and outputs through our API.

Google

Google API usages offer a free quoto currecntly, and google will use your data if you are using this free quota. But if you enable payemnt via a paid GCP project, your data will not be used for traning is what is thier in Terms Page

As per OpenA Terms Docs on 5th July 2024, they state,

When you're using Paid Services, Google doesn't use your prompts (including associated system instructions, cached content, and files such as images, videos, or documents) or responses to improve our products, and will process your prompts and responses in accordance with the Data Processing Addendum for Products Where Google is a Data Processor. This data may be stored transiently or cached in any country in which Google or its agents maintain facilities..

If you want to know more, please visit the LLM providers terms pages.