Text to speech

Text to speech is the voice layer when the picture already exists, or when the picture will be assembled later. Faceless channels, translated Shorts, and cloned-narrator TikToks all hit a credit pool before they hit a timeline. This matrix maps the eight tools already verified on this site to web TTS, a documented API, and an official MCP endpoint. Murf has a tool page, verified 2026-10-17, and stays off this first-tier matrix. PlayHT and Speechify are still not listed. play.ht closed the connection and speechify.com returned a bot challenge on 2026-10-17, so those prices are not observed.

How to read this table

• Web is observed when the tool page documents in-product speech generation or included AI voices for creators. A handoff that says "use ElevenLabs instead" stays not_observed for that row.
• API quotes the developer URL from the tool frontmatter when present. Kling's text-to-audio line is an API surface note, not a claim that the membership app is a TTS studio.
• MCP lists only official URLs printed on tool pages. ElevenLabs MCP is explicitly not the full dubbing or music API.
• Price copies plan stickers verified 2026-10-16. ElevenLabs Starter's first-month $1 label was dated until Oct 18 on that check; budget the renewal $6.
• Verified matches the tool-page date. Re-check commercial-license text yourself; this directory did not open ElevenLabs contract prose.

Selection tips

• Default narration stack for short-form without a face: ElevenLabs on voice and publish, then CapCut only if you still need type or a picture edit.
• Brand publishes should start at ElevenLabs Starter or higher on the plan card that prints a commercial-license line. Free is framed as a draft pool on the tool page.
• Avatar products (HeyGen, Creatify) include voices because the product is a person on screen. Prefer them when the face is required; prefer ElevenLabs when the track is the product.
• CapCut can meter premium text-to-speech through AI credits on paid plans, per the help note on the CapCut tool page. It is not the default voice studio on this directory.
• Kling's developer list includes text-to-audio. Treat that as an API line item until a membership TTS workflow is filed separately.

Related workflows and models

• Workflows: voice-publish, faceless, ai-avatar, ugc-ads.
• Models: video models here do not replace a licensed narrator. Pair Kling or Seedance b-roll with ElevenLabs audio.

What creators usually get wrong

The common mistake is treating every product that can speak as interchangeable. ElevenLabs meters a shared credit pool across speech, sound, and music on the compare extract. HeyGen and Creatify speak because a face or an ad actor is on screen. CapCut can spend credits on premium text-to-speech inside an editor, which is a finish, not a narrator studio. Kling's text-to-audio line lives on the developer list beside video modes, so it is an API planning note until a membership TTS walkthrough is filed on this site.

Another mistake is budgeting the ElevenLabs first-month teaser as the forever price. The Starter card checked on 2026-10-16 labeled a $1 first month until Oct 18 against a $6 renewal. Creator's first month is $11 against $22. This matrix copies those stickers and stops. It does not invent a per-character dollar from an older blog post, and it does not open the commercial-license contract text.

When the publish target is YouTube narration with a clone of your own voice, keep the disclosure distinction already used on the voice-publish workflow: an own-voice narration example on YouTube help materials is not the same upload as a realistic AI face. TikTok may still ask for a label on a significant AI modification. Those sentences are workflow policy pointers, not new legal advice invented for this matrix.

FAQ

Is ElevenLabs Free enough to publish under a brand?

The tool page keeps Free as a $0 draft plan without the commercial-license line that Starter prints. This directory did not open the license contract. Read the card and the terms before you ship.

Does HeyGen accept an ElevenLabs file?

Audio input appeared as a comparison-table row on the HeyGen extract; which plans include it was not observed as a clear yes or no. Do not add ElevenLabs dollars into a HeyGen starter sum until your account shows the import.

Why is Runway not observed for web TTS?

The Runway tool page tells creators to replace voice in ElevenLabs when Runway audio is not the licensed track. That is a handoff, not a TTS product claim.

Does CapCut Pro include unlimited TTS?

No unlimited claim was filed. Credits pay for advanced AI use such as premium text-to-speech; credits do not buy the Pro subscription itself.

Where is Murf or PlayHT?

Murf has a tool page, verified 2026-10-17, and is not a row in this first-tier matrix. PlayHT and Speechify are still not listed. play.ht closed the connection and speechify.com returned a bot challenge on 2026-10-17, so those prices are not observed.