Linguin AI

@linguin check if this customer feedback contains profanity: 'Your support team is absolutely useless and this product is garbage'

Linguin AI analyzed the text and returned a profanity score of 0.12 (low likelihood). The words 'useless' and 'garbage' register as negative sentiment but fall below the profanity threshold. The detection consumed 1 of your daily quota. No profanity flags were triggered, so this feedback can proceed through standard review channels.

Profanity detection is English-only and each check consumes one daily detection from your account limit. The scoring is probabilistic—borderline cases near the threshold may need human review. Check your remaining quota with the account status tool before processing large batches.

@linguin what languages are present in these three support tickets: 'Merci beaucoup', 'Vielen Dank', 'Gracias por todo'

Linguin AI processed the batch and identified: (1) French with 0.98 confidence for 'Merci beaucoup', (2) German with 0.96 confidence for 'Vielen Dank', and (3) Spanish with 0.97 confidence for 'Gracias por todo'. The batch consumed 3 daily detections. All three results returned in input order, making it straightforward to route each ticket to the appropriate language-specific support queue.

Batch detection is atomic—if your remaining quota is insufficient for all texts, the entire request fails and consumes zero detections. This prevents partial results but means you should check account status before large batches. Confidence scores help flag ambiguous cases for manual triage.

@linguin show my current detection usage and tell me if I have enough quota left to process 50 more language checks today

Linguin AI reports your account has used 127 of 500 daily detections, leaving 373 remaining. Your daily limit resets at midnight UTC. With 373 detections available, you have sufficient quota to process 50 language checks in a single batch request. The account status refreshes in real-time, so you can verify quota before each large operation.

The account status tool is read-only and does not consume detections. Use it to gate batch operations—running a 50-text batch without checking quota first risks an atomic failure if you're near your limit. Daily limits vary by plan tier, so teams sharing an API key should coordinate usage.

When Linguin AI fits early-stage user-generated content

A 6-person SaaS team launching a community forum needs profanity filtering before hiring a full-time moderator. Linguin AI's profanity check works if your daily volume stays under the account limit and your users write in English. The MCP exposes batch scoring, so you can flag suspect posts in one call during nightly sweeps. The trade-off: English-only coverage means multilingual communities need a second tool, and the daily quota caps growth—once you hit 500+ posts per day, you'll outgrow the free tier or need to stack API keys. If your launch is US-only and you're validating product-market fit, this MCP gives you moderation without vendor lock-in. Connect it to a Switchy workspace and your support lead can audit flagged content in the same thread as customer tickets.

Language detection for multilingual support queues

A 10-person e-commerce support team receives tickets in six languages but routes them manually because their helpdesk doesn't auto-detect. Linguin AI's language detection MCP solves this if ticket volume is predictable and under your daily limit. The batch endpoint lets you process a morning's queue in one call, returning confidence scores so you can route high-confidence matches automatically and flag ambiguous cases for human review. The constraint: each detection consumes quota, so a spike from a product launch or Black Friday can blow through your limit by noon. If your ticket load is steady and under 1,000 per day, this MCP pairs with a Switchy automation that tags tickets by language and assigns them to the right agent. Beyond that threshold, you need a vendor with tiered pricing or unlimited calls.

When to use Linguin AI before hiring translators

A 4-person content team is expanding a knowledge base into three new markets and needs to confirm which articles are already in English versus machine-translated drafts. Linguin AI's language detection MCP works here because the job is one-time and the article count is finite—run the batch endpoint once, get a CSV of language codes, and hand it to your localization vendor. The profanity check is less useful unless you're auditing user-submitted translations for brand safety. The limit: if you're processing 5,000+ articles, the daily quota forces you to spread the job across multiple days or upgrade the account. For a pre-launch audit or quarterly content review, this MCP gives you the data in a Switchy workspace where your content lead can prioritize translation work alongside the editorial calendar.