Fair Use Policy

1. Overview

This Fair Use Policy applies to all Intentgine subscription plans. It exists to ensure a reliable, high-quality service for all users. Intentgine is operated by Neural Core LTD, a company registered in England and Wales.

Each plan includes a monthly request allowance and resource limits. These limits are designed to support normal, expected usage patterns for each tier.

2. Plan Limits

The following limits apply per billing cycle:

New accounts receive 1,000 free trial requests before subscribing.

3. What Counts as a Request

Every API call counts as at least one request against your allowance, regardless of whether the response is served from cache or requires LLM processing.

Some endpoints consume more than one request per call. For example, /v1/resolve-quick costs 2 requests per call as it always performs a fresh LLM computation without caching. Similarly, /v1/resolve-respond costs 2 requests per call as it performs both tool resolution and conversational response generation. The /v1/classify-respond endpoint costs 2 requests for single-intent or 4 requests for multi-intent with extraction.

For batch endpoints such as /v1/classify-batch, you are charged 1 request per unique item in the batch. Cache hits are billed at half rate (1 request per 2 cached items, rounded up). Cache misses are deduplicated via embedding similarity — items with 95% or greater cosine similarity are treated as semantically equivalent and only counted once.

The requests_used field in each API response tells you exactly how many requests were consumed.

4. What Happens When You Reach a Limit

When you reach your monthly request allowance or a resource limit:

• API requests that exceed your allowance will return an error response
• Existing resources (apps, banks, toolsets) remain accessible but you cannot create new ones beyond your limit
• You can upgrade your plan at any time to increase your limits
• Request allowances reset at the start of each billing cycle

5. AI Output Accuracy

Intentgine uses large language models and AI to process your requests. AI outputs are probabilistic and may vary between requests. You should be aware that:

• Results may be inaccurate, incomplete, or inconsistent, particularly for ambiguous or novel inputs
• Classification confidence scores are estimates, not guarantees of correctness
• We may change the underlying model providers or versions to improve quality, which may affect output behaviour
• Caching and deduplication mean that repeated or similar inputs may return previously computed results rather than fresh AI processing

We recommend implementing validation, fallback logic, and human review where AI outputs inform consequential decisions. See our Terms of Service for full warranty disclaimers and limitation of liability.

6. Prohibited Usage Patterns

The following are considered unfair use regardless of plan:

• Automated request flooding or denial-of-service patterns
• Creating multiple accounts to circumvent plan limits
• Sharing API keys with third parties not covered by your account
• Using the service in a way that degrades performance for other users
• Systematically scraping or extracting model outputs for the purpose of building competing services

7. Enforcement

If we detect usage that violates this policy, we may:

• Contact you to discuss your usage and suggest an appropriate plan
• Temporarily throttle or rate-limit your API access
• Suspend your account if the issue is not resolved

We will always attempt to contact you before taking action, except in cases of clear abuse or security threats.

8. Changes to This Policy

We may update this policy and the associated plan limits from time to time. We will notify you of material changes via email or through the service. Current limits are always reflected on our limits page.

9. Contact

If you have questions about this policy or need to discuss your usage, contact us at support@intentgine.com.