ElevenLabs

Published September 3, 2026 · This page indexes everything this site has verified about ElevenLabs and links to where each finding is set out in full. It is not a review — the review is. No listening test has been run, so nothing here says how ElevenLabs sounds.

On this page

What it costs

Plans run Free $0 (10,000 credits), Starter $6, Creator $22 (first month $11), Pro $99, Scale $299 and Business $990 a month, with annual billing charging for ten months out of twelve — equivalent to $5 Starter, $18.33 Creator, $82.50 Pro, $249.17 Scale and $825 Business; API list prices are $0.05 per 1,000 characters on Flash/Turbo and $0.10 on Multilingual v2/v3, and the rate is identical at every tier, so higher plans buy allowance rather than a better unit price. Those figures were verified August 11, 2026 on the review and re-read September 1, 2026 on the comparison pages. Dividing plan price by included credits gives a plan-effective $0.165–$0.200 per 1,000 Multilingual characters (computed August 11, 2026), which pricing comparison restates as $165.00 to $200.00 per million from rates read September 1, 2026 — the highest figures on that page, against a published market floor of $4.00 per million — while dubbing is metered separately per source audio minute at $0.33 watermarked, $0.50 unwatermarked and $2.20 on v2 (read September 2, 2026). Every one of the subscription figures carries the standing caveat that the vendor’s two official pricing pages disagree about what a plan includes.

Every rate above is set against every other provider’s, normalized to one unit, in our pricing comparison.

Where ElevenLabs’s own material disagrees with itself

These are contradictions inside the vendor’s published documentation, not disagreements between us and the vendor. We report both readings rather than choosing one, and each links to the page where we set it out in full. We have published 17 in total; the 10 with the most bearing on a decision are below, and the rest sit in the pages linked underneath.

• ElevenLabs publishes two official pricing pages that describe plan inclusions on different bases and disagree by roughly 1.65× to 2× on what a subscription includes, and the vendor never reconciles them; the site attributes both figures to their sources, refuses to average them, and advises budgeting against the more conservative credits basis. Checked August 11, 2026. Where we published this.
• The mismatch runs the whole plan ladder: the main pricing page sells Starter as 30,000 credits against the API page’s 60,000 characters, Creator 121,000 against 220,000, Pro 600,000 against 990,000, Scale 1,800,000 against 2,990,000 and Business 6,000,000 against 9,900,000 — only the free tier matches, at 10,000 either way. Because ElevenLabs itself states that text-to-speech costs one credit per character, both sets of figures cannot be true, and the site concludes that no reliable effective cost per character can be computed for any paid plan while the contradiction stands. Checked September 1, 2026. Where we published this.
• The pricing contradiction is listed among the documented reasons to leave ElevenLabs: its two official pricing pages state incompatible allowances for every paid tier — Starter 30k credits against 60,000 characters, Business 6M against 9,900,000. Both are official; neither is reconciled. Checked September 1, 2026. Where we published this.
• One ElevenLabs page carries two incompatible language totals: the heading on its text-to-speech page says "over 70 languages" while the FAQ lower on that same page says "32+ languages across our model lineup". The two enumerated lists disagree as well — the site counted 74 entries in the documentation and 76 names in the marketing list, and endorses neither. Checked August 11, 2026. Where we published this.
• In the site’s own recount of every vendor’s language claims, ElevenLabs is one of two providers publishing two headline figures on a single page — "over 70" and "32+" — against the site’s counts of 74 entries in docs and 76 names in marketing: two headline figures that cannot both be right, and two lists that disagree with each other. Its enumerated list is therefore classed as only partly readable. Read September 2, 2026. Where we published this.
• ElevenLabs publishes three different latency figures for the same Flash v2.5 model family, each naming a different metric: ~75 ms for model inference, 100–150 ms TTFB across three regions (rising to 150–200 ms across two more), and sub-500 ms end-to-end. The site notes the three need not conflict, because each measures a different quantity — but only one is the number a user would experience, and none of them is verified by this site. Read September 2, 2026. Where we published this.
• ElevenLabs’ own documentation undercuts its own headline number: it states that model inference latency "is an internal measurement, excluding network round-trips and application overhead", while time-to-first-audio "is almost always the number that matters for user experience, and it is always larger — often substantially larger — than model inference latency alone." No TTFA figure is attached to it anywhere on that page. Read September 2, 2026. Where we published this.
• The dagger on ElevenLabs’ "~75ms†" headline resolves to a scope note rather than a definition — "† Excluding application & network latency" — but two of the three places its ~280 ms conversational-model figure appears carry no dagger at all, so the exclusion travels with the number only some of the time. Read September 2, 2026. Where we published this.
• ElevenLabs states its dubbing language coverage three incompatible ways: the documentation says "90+", while the pricing cards say 29 for Dubbing v1 and 92 for Dubbing v2. The site counted the documentation’s own tables and found 86 rows for v1 and 94 for v2, matching none of the published figures. Read September 2, 2026. Where we published this.
• ElevenLabs publishes two different dubbing upload limits — 1 GB on one page and 2 GB on another — and the site records the conflict as unresolved rather than choosing between them. Read September 2, 2026. Where we published this.

Dated and time-sensitive

Facts with a clock on them — model sunsets, policy revisions, market status. These are the entries most likely to be out of date first, which is why they carry their dates.

• 2026-03-31. ElevenLabs’ Terms of Service (non-EEA variant) carries a last-updated date of March 31, 2026. Clause 1(c) sets commercial use by tier — free users non-commercial only, paid users commercial — and §4(b) governs cloning consent while §4(c) leaves output rights with the user. All ToS citations on this site are to the non-EEA document; the EEA variant was not reviewed. Source on this site.
• 2026-08-17. ElevenLabs’ Prohibited Use Policy now displays "Last Updated 17 August 2026". Because Terms clause 1(c) incorporates the policy into the commercial-use rule, its version date is material, and the site advises that anything concluded from a reading before that date is worth re-reading. Source on this site.
• 2026-09-01. Correction logged on the review’s change log: the page had cited the Prohibited Use Policy as "updated Sep 3, 2025"; re-checked September 1, 2026, the document displays "Last Updated 17 August 2026" — six days after the review’s original verification date. The citation was corrected in both the body and the sources table. The site states it has not yet re-read the full policy text against its August summary; that re-read is scheduled. All other facts on the page remain as verified August 11, 2026. Source on this site.
• 2026-08-11. Three ElevenLabs models are recorded as deprecated and excluded from the review’s current model line: eleven_turbo_v2_5, eleven_turbo_v2 and eleven_multilingual_v1. The current line is Eleven v3, Multilingual v2 (the API default), Flash v2.5 and Flash v2. Source on this site.
• 2026-09-02. Dubbing lifecycle: Dubbing Studio runs only on the v1 model, which the vendor states "is in maintenance mode and receives critical bug fixes only", while Dubbing v2 is self-labelled "currently in alpha" and produces dubs "completely automatic without any option to edit the content". Lip-sync is explicitly not offered and real-time dubbing is "not currently available". Source on this site.
• 2026-09-02. The optimize_streaming_latency parameter is documented as deprecated on both the convert and stream endpoints of ElevenLabs’ API reference. Source on this site.
• 2026-09-02. Transport limits re-read: the TTS WebSocket does not support the eleven_v3 model, and multi-context WebSocket contexts time out after a default 20 seconds of inactivity. Source on this site.
• 2026-09-02. Per-plan concurrency re-verified: Multilingual v2 allows 2 concurrent requests on Free, 3 Starter, 5 Creator, 10 Pro, 15 Scale and 15 Business, with Flash double throughout; Enterprise is published as "Elevated" rather than a number. Exceeding the limit queues rather than fails, which the vendor documents as adding about 50 ms of latency. Source on this site.

Every page on this site about ElevenLabs

• Full review. ElevenLabs Review 2026: Pricing, Licensing and Voice Cloning, Docs-Verified — the complete documentation-based assessment, with sources and dates against every claim.
• Alternatives. ElevenLabs alternatives — organized by the reason you would leave, not as a ranked list.
• Head to head. ElevenLabs vs Google Cloud, ElevenLabs vs Murf AI.
• In our buying guides. Best AI Voice Generators in 2026: 10 Providers Compared, Best AI Dubbing Software 2026: Only Four Vendors Actually Sell It, Best Free AI Voice Generators 2026: What “Free” Actually Buys You, Most Realistic AI Voices: What Every Vendor Charges for the Claim, Best Multilingual AI Voice Generators: Every Count, Recounted, Real-Time AI Voice: Why the Latency Numbers Cannot Be Compared, Best AI Voice Cloning Software 2026: Price, Access and Consent Compared.
• In our comparisons of published data. AI Voice Generator Pricing Comparison: Every Rate Normalized, Best Text-to-Speech APIs: Limits, Transports and Auth Compared.
• In our explainers. Can You Use AI Voices Commercially? What Ten Vendors Actually Say, How AI Voice Cloning Works: What Ten Vendors Require of You, How AI Voice Generators Work, According to the Vendors Themselves.
• By what you are making. AI Voice for Audiobooks: Will Chapter 20 Match Chapter 1?, AI Voice for Business: What Survives a Procurement Review, AI Voice for eLearning: Courses Change, and the Audio Must Follow, AI Voice for Podcasts: Only Three Providers Do Two Speakers, AI Voice for YouTube: What You Must Disclose Before You Publish.

What we could not verify

Honesty about gaps beats a page that looks complete. Standing limits on everything above, as of September 3, 2026:

• How ElevenLabs sounds. Our listening benchmark is designed but not running. Nothing on this page or on any page it links to ranks ElevenLabs on audio quality, naturalness or speed.
• Anything ElevenLabs does not publish. Every page indexed here is built from the vendor’s own documentation. Where that documentation is silent, we record the silence rather than estimating around it.
• Whether a documented feature works as documented. We read documentation; we did not create an account or make an API call.
• Currency. Each linked page carries its own check date. This index inherits them and does not re-verify on load.

Change log

September 3, 2026 — First publication. Assembled from pages already published on this site; no new vendor claim was introduced by this page.

Published September 3, 2026 · This page shows no audio and reports no listening results, because our benchmark has not run. Corrections are recorded with their date on the corrections page. Independence note: this page contains no affiliate links, and no vendor paid for placement or influenced the order — see how we make money and our editorial policy.