AI USAGE PRICING AND TERMS

Introduction

This document outlines the pricing and terms of use for AI operations within Metance. Metance Tokens are used as the currency for these operations. User-initiated AI operations generally require Metance Tokens. The AI is also inherently used by the system for categorization, processing, and analyzing content. The cost of these kinds of operations are included in the plan, and no extra charges will apply. This document aims to detail the usage of Metance Tokens, the approximate cost of operations, settings, and management of usage. It is important to mention that not all plans include AI functions. These terms apply only to subscription plans that include AI functionality.

Token Economy

As of January 1, 2024, the cost for 1000 Metance Tokens is approximately $1. Tokens are allocated to each workspace based on the number of users, with the total allocation being 1000 tokens multiplied by the user count. These tokens form a shared pool for the workspace, rather than being individually allocated to each user. Usage beyond the free allowance is billed at the end of each billing cycle.

Overage Preferences Management

Admins can adjust overage preferences in the settings under the AI section. Options include allowing indefinite overage, setting no overage, or establishing controlled overage limits. Once these limits are reached, AI operations will not be permitted for users. AI operations will function if the AI setting is activated in the settings. It is also possible to toggle each sub-setting on or off to enable or restrict access to specific features. The Content-Related section includes features where users can select to use AI assistance for creating content, generating images, or receiving suggestions. Options under the administrative section generally involve automatic functions, such as content analysis, question generation, and topic associations. Activating these options can lead to better content classification and easier discovery. Depending on each company’s needs, some of these options can be activated, while others may be left off.

AI Operation Costs

Image Operations:

• Image Generation: Around 100 Tokens for a set of 3 alternative images. Image generations are initiated by the users wherever it is available.

• Image Analysis: Image analysis costs approximately 20 Tokens per image. This includes extracting descriptions, dominant colors, sharpness, themes, and other metadata. The analysis can be initiated by users. Additionally, if the relevant setting is enabled, it can run automatically whenever an image is added to the workspace.

Content Analysis:

• General Analysis: Approximately 10 Tokens per content.

• Descriptive Analysis: 5 to 30 Tokens, depending on workspace content density. This analysis includes checking content meaningfulness, information relevance, department association, sensitivity check, feasibility scoring, and classification.

Text Operations:

• Basic Operations: These are operations like ‘Continue Writing’, ‘Spell Checking’ and so on. They cost about 1 Token for every 500 words processed. They are initiated by the users while creating content in the editor.
• Content-Based Functions: The cost is an average of 2 Tokens for every 500 words, varying based on the complexity and processing needs of the operation. Content-based functions are initiated by users, and the results may be specific to them. Most of the time, these are not stored in the workspace. This can lead to different team members repeatedly running the same functions, which may become costly. We advise saving the results of these functions as supplements to avoid extra costs.

Classification Operations:

• Attribute Suggestions: The cost varies based on content length, total number of ontent types, and topics in the workspace. For new workspaces, the average cost is around 5 Tokens, while in more established workspaces, it averages about 10 Tokens.

• Auto Topic Association: Token usage varies with the number of topics in the workspace and content length.

Details about token consumption

Operations on larger content types like Articles and Reports incur higher costs due to their complexity and size. Costs for chatbot widget may differ based on the GPT model selected. For instance, using GPT-4 consumes 10 times as many tokens as GPT-3.5. Additionally, size of company brief in the settings can slightly affect Chatbot Widget token costs.

Token Validity and Rollover

Free Metance Tokens are strictly tied to the current billing period. This means that these tokens are only available for use within the specific timeframe of the current billing cycle. Once this cycle concludes, any tokens that have not been utilized will no longer be valid. They do not accumulate or roll over into the next billing period.

Accuracy of AI Functions

We continuously refine our algorithms and models to ensure the utmost precision all AI-related tasks. But it’s important to recognize that AI operations, by their nature, may not always be perfect. The accuracy of AI outputs can vary based on the complexity of the task, the quality of input data, and the specific AI model being used. Users should be aware that the token cost is based on the operation performed, not the accuracy of the result.

Overages Cap

To protect users from unexpectedly high costs, overages are capped at 25,000 Metance Tokens, regardless of the overage action selected. If this cap is reached, only user-initiated AI functions will be limited. System-level AI functions will continue to operate.

Pricing and Terms Adjustments

Metance reserves the right to adjust token pricing and terms of use. Users will be notified in advance of any significant changes to ensure transparency and allow for planning.

Support and Contact

For any queries, clarifications, or assistance regarding these terms and pricing, users are encouraged to contact the Metance support team. Our team is dedicated to providing comprehensive support to ensure an optimal user experience.