LMNT

@LMNT list all available voices and show me which ones are best for professional narration

Here are the voices available in your LMNT account: **Professional Narration (recommended):** - **Morgan** (neutral, clear articulation, ideal for tutorials) - **Alex** (warm, authoritative, suited for corporate content) - **Jordan** (conversational, engaging, works well for explainers) **Other voices:** - Casey (casual, friendly) - Riley (energetic, youthful) - Sam (technical, precise) This is an illustrative example showing how LMNT would return voice metadata. Your actual account may have custom voices you've created or different preset options depending on your subscription tier.

This example demonstrates the voice discovery workflow. Before synthesizing speech, you'll need to know which voice IDs are available. Note that LMNT's voice library varies by account type, and custom voice creation may require additional credits or subscription features.

@LMNT synthesize this script using the Morgan voice: 'Welcome to today's product demo. We'll cover three key features in the next five minutes.'

Speech synthesis complete. I've generated audio using the Morgan voice for your script. **Audio details:** - Duration: ~8 seconds - Voice: Morgan (neutral/professional) - Format: MP3, 24kHz sample rate - File size: ~192 KB The audio file is ready for download or playback. This illustrative example shows the synthesis confirmation you'd receive. The actual output includes a playable audio URL that expires after a set period (typically 24 hours).

This write action consumes synthesis credits from your LMNT account quota. Each synthesis request counts against your monthly limit, so be mindful when generating multiple takes. The API returns a temporary URL to the audio file—download it promptly if you need permanent storage.

@LMNT create a new custom voice called 'BrandVoice' and check my account's remaining synthesis credits

I've initiated voice creation for 'BrandVoice' and retrieved your account status: **Voice Creation:** Custom voice 'BrandVoice' has been created with ID `voice_abc123xyz`. You can now use this voice for synthesis by referencing this ID. Note that training a custom voice typically requires uploading audio samples separately through LMNT's interface. **Account Status:** - Remaining synthesis credits: 8,450 characters - Subscription tier: Professional - Custom voices allowed: 5 (you've now used 2) This example illustrates how LMNT confirms resource creation and reports quota usage.

Custom voice creation is a powerful feature but requires careful setup. The MCP can create the voice entity, but you'll typically need to provide training audio samples through LMNT's web interface or separate upload endpoint. Monitor your character quota closely—synthesis limits reset monthly based on your subscription.

When LMNT keeps your demo videos on-brand

A 6-person SaaS startup records weekly product walkthroughs for their knowledge base. They need the same voice across 40+ videos, but their narrator quit. LMNT wins here: create one custom voice, then synthesize every script through Get Synthesize Speech without re-recording. The voice management tools (Get Voice Info, Update Voice Info) let the marketing lead tweak tone as the product matures. The threshold: if you're doing real-time conversational AI or need sub-200ms latency, LMNT's batch synthesis model won't cut it. But for async content where brand voice matters more than speed, this is the call. Set up the API key once, point your content pipeline at the voice ID, and stop scheduling studio time.

LMNT for small-team phone support automation

A 3-person support team at a B2B tool wants to automate their after-hours phone tree without sounding robotic. They have 12 menu prompts and 8 hold messages that change quarterly. LMNT is overkill if you're just using stock voices—Twilio's built-in TTS is cheaper. But if brand voice is part of your support experience (think: a founder's voice welcoming callers), the Create Voice and Get Voices List tools justify the setup. You'll synthesize each prompt once per quarter, cache the audio, and serve it through your IVR. The trade-off: LMNT charges per character synthesized, so if your scripts are over 10k characters monthly, watch your usage. For boutique support where voice = brand, this is the right tool.

When LMNT speeds up podcast pre-production

A 2-person media team produces a weekly interview podcast. Before recording, they want to hear their intro and outro scripts read aloud to catch pacing issues. LMNT's Get Synthesize Speech tool turns a 300-word script into audio in seconds, letting them iterate on copy before the host records. The Get Voices List tool helps them A/B test narrator styles without hiring voice actors. This isn't for the final published episode—human narration still wins on emotional range—but for draft review, LMNT cuts the feedback loop from days to minutes. The boundary: if your scripts are under 100 words or you're already recording everything live, the API overhead isn't worth it. For teams writing long-form audio content, this is the pre-production shortcut you didn't know you needed.