DEV Community: Jaden Vo

I built 2 free tools for creators who use AI voiceovers (no signup)

Jaden Vo — Wed, 21 Oct 2026 08:54:45 +0000

I do a lot of narration work, and for a long time I estimated two things by gut feel: how long a voiceover would run, and what the text-to-speech bill would be. Both were guesswork, and both were wrong often enough to annoy me. So I built two small tools and put them online for free.

What they do

Script Reading Time & Voiceover Length Calculator: paste a script and it shows words, characters (with and without spaces), and narration length at 130, 150 and 170 WPM. It also shows roughly how many characters a TTS tool will bill.

AI Voice (TTS) Pricing Comparison: enter N characters (or minutes) per month and it estimates the monthly cost across ElevenLabs, Murf API, WellSaid, Google Cloud TTS and Azure, cheapest first.

There is more in the hub: https://ai.infinityme.dev/tools/

How it's built

It's a static Next.js export, and everything runs client-side. The text you paste never leaves your browser, and I don't track it. No signup, no accounts, no affiliate links in these tools.

Lessons learned

Pricing pages change constantly. Hardcoding numbers in components went stale fast. Now every price lives in one data file, each entry with a "checked" date, and the page shows that date. Updating means editing one file.
Characters vs minutes billing confuses people. Some providers bill per character, others sell plans in minutes or credits. I had to pick a conversion (about 1,000 characters per minute) and state it clearly, otherwise comparisons look misleading.
Estimates need honest margins. Real recordings include pauses and retakes, and regenerating a line costs characters again. The tool gives a number, but you should leave some slack.

Feedback welcome

If something looks off, or a price is outdated, tell me. Which TTS provider should I add next? Tell me in the comments, and tell me what else would make these useful for your workflow.

Estimating ElevenLabs Costs for an Automated Video Pipeline (2026)

Jaden Vo — Tue, 20 Oct 2026 17:08:28 +0000

If you generate voiceovers programmatically, "which plan?" is really a capacity-planning question. Here's how I size ElevenLabs plans, using the official pricing page (checked October 20, 2026).

Units

1 character ≈ 1 credit (standard multilingual TTS). Flash/Turbo models are discounted for API use (0.5–1 credit/char).
~1,000 characters ≈ 1 minute of narration.
One credit pool is shared by TTS, STT (≈330 credits/min), sound effects (≈200/generation), voice changer and isolator (≈1,000/min) and dubbing (2,000–10,000/min, depending on mode).

Quick estimator

// rough monthly credit estimate
const est = ({ videosPerMonth, minutesPerVideo, retakeFactor = 1.3, sfxPerVideo = 0 }) =>
  Math.ceil(videosPerMonth * (minutesPerVideo * 1000 * retakeFactor + sfxPerVideo * 200));

est({ videosPerMonth: 4, minutesPerVideo: 10, sfxPerVideo: 5 }); // ≈ 56,000 → Creator (121k)
est({ videosPerMonth: 20, minutesPerVideo: 1 });                 // ≈ 26,000 → Starter (30k)

Plans (monthly / annual-equivalent)

Plan	Monthly	Annual/mo	Credits
Starter	$6	$5	30k
Creator	$22 (first month $11)	$18.33	121k
Pro	$99	$82.50	600k
Scale	$299	$249.17	1.8M

Things that affect your bill

Rollover: unused credits on paid plans roll over for up to 2 months (max balance 3× quota). Downgrading or cancelling forfeits them, so drain the balance before you switch plans.
Per-generation charging: you pay per request, not per download, so cache aggressively and regenerate only changed chunks.
Commercial licence: the Free tier is non-commercial. Any monetized output needs Starter or above.
Audio format: 44.1 kHz PCM via the API starts on Pro. If your pipeline needs uncompressed output, factor that in.

Recommendation

Prototype on Free, run production on Starter or Creator depending on your estimator output, and switch to annual billing once your usage has been stable for 2–3 months.

Full plan-by-plan breakdown: ElevenLabs Pricing Explained (2026).

ElevenLabs vs Murf AI for Developers (2026): API, Pricing and a TTS Pipeline for Video

Jaden Vo — Tue, 20 Oct 2026 11:33:45 +0000

I automate voiceovers for video content, so I compared ElevenLabs and Murf AI the way a developer would: pricing model, API, output formats and how each fits into a script-to-audio pipeline. All prices are from the official pricing pages, checked October 20, 2026.

Pricing models are different, and that changes your architecture

ElevenLabs sells subscription plans with a monthly credit pool shared across every product (TTS, speech-to-text, sound effects, dubbing). For its standard multilingual TTS models, 1 character ≈ 1 credit. The pricing FAQ says the Flash/Turbo models get discounted API pricing, at 0.5–1 credit per character.

Plan	$/month	Credits	Notes
Free	0	10k	Non-commercial
Starter	6	30k	Commercial licence, Instant Voice Cloning
Creator	22	121k	Professional Voice Cloning
Pro	99	600k	44.1 kHz PCM output via API, 192 kbps

Unused paid credits roll over for up to 2 months (capped at 3× your quota) while you stay subscribed.

Murf prices its API as pay-as-you-go: $0.01 / 1k chars for Falcon (built for real-time voice agents) and $0.03 / 1k chars for studio-quality TTS. You get $10 of free credit, and purchased credit doesn't expire. The pricing page lists up to 5 concurrent requests on standard plans.

What this means in practice: for spiky, high-volume batch jobs, Murf's pay-as-you-go model is easy to budget. For a steady weekly pipeline where you also want cloning, sound effects or dubbing, an ElevenLabs subscription gives you more per dollar of tooling.

A minimal pipeline pattern

Whichever vendor you use, the shape is the same:

script.md
  → split by paragraph (keeps regenerations cheap: you pay per generation)
  → normalize text (numbers, acronyms, phonetic names)
  → TTS per chunk (cache by hash of text + voice + settings)
  → concat + loudness-normalize (ffmpeg loudnorm)
  → export WAV/MP3 for the editor

Two tips that save real money:

Hash-based caching. Key each chunk on sha256(text + voice_id + settings). Fix one typo and only that chunk regenerates.
Chunk by paragraph, not by sentence. Sentence-level chunks lose prosody across boundaries, while whole-script generation makes every fix expensive.

Quality and features

Expressiveness: ElevenLabs is generally more natural for narration and emotion. Murf sounds like a clean, neutral presenter.
Cloning: ElevenLabs offers Instant from Starter and Professional from Creator. Murf lists Professional Voice Cloning under Enterprise.
Languages: ElevenLabs' plan table lists 74 TTS languages. Murf lists 150+ voices across 35+ languages for its API.
Studio/timeline: Murf's web Studio is better for syncing narration to slides. ElevenLabs' Studio targets long-form audio.

Verdict

Building a content pipeline (YouTube, podcasts, audiobooks) and want the best voices and cloning: ElevenLabs.
Building a voice agent or high-volume TTS feature on a pay-as-you-go budget: benchmark Murf's Falcon too.

The full comparison (tables, FAQ, non-dev considerations) is here: ElevenLabs vs Murf AI (2026).