- حمل الفيلم - قول لكلود يحولك الفيلم لأوديو فايل - خد الفايل حوله text على elevenlabs speech to text - إعمله export ك srt هيطلعلك فايل ترجمة بالtimestamps بس باللغة الأصلية - خد الفايل وإديه لكلود وقوله ترجمه وركبه عالفيلم (عملتها كذا مرة بس مكانتش أفلام)
brandSteady
ElevenLabs on X
The largest group, 46 of 124 people, talks about finished work — voiceovers, short films, games, receptionists — with the voices praised as hard to tell from a real person. Complaints centre on the meter: per-character bills, credits that drain, and slow billing support.
First written 27 Oct · what changed · from 138 posts by 124 people
Tracking posts that say “elevenlabs”
Track this for your marketPosts, last 24h
57
−42%vs. previous 24h
Posts, last 7 days
392
271 in the last 72h
Velocity
56/day
7-day average
Momentum
+2%
Last 24h vs. a normal day this week
Posts per day, last 30 days
as of Nov 3, 11:01 UTCWhat changed
as of 27 Oct · checked once a monthFirst written 27 Oct. When what people say changes, it shows here.
Tone· 110 posts from 96 people· written 27 Oct
How do people feel about ElevenLabs?
Talk is built around work people finished: 46 of the section's 96 people describe a voiceover, short film, game or tutorial it voiced, far ahead of any other theme. Plain enthusiasm (18 people) and praise for how real and emotional the voices sound (18 people) sit level behind it.
People name the job rather than the model: mini documentaries, product tutorial series, short films, in-game radio ads, a browser game that got two hours of audio drama in five minutes. One built a whole tutorial series through the MCP and called it "the highest leverage thing" they had done; another's first short film credits ElevenLabs for sound and music.
Voice cloning carries much of this: users say a three-minute recording makes a natural clone, and that writing a line the way you'd say it, plus an emotion tag, gets more out of v4. Praise for the voices themselves is a separate group of the same size — natural, expressive, "hard to tell apart from a real person's", with Japanese and other non-English output named as improved.
A smaller set stays on speed and boards: v4 Turbo under 150 ms to first speech, Eleven Realtime Max at 0.57 s and 98.8% on Big Bench Audio, top spots on five Voice Arena language boards.
Dissent is real but smaller: each model has tradeoffs (V3 for expressive fiction, V2 for non-fiction), some output "lacks human touch", and one reader calls the board wins statistical ties by Voice Arena's own rule.
- Happy with what they make49
- Loves it21
- Sounds real and emotional18
- Reservations and criticism13
- Fast and top of the benchmarks9
See 5 of 110 postsHide the posts
What if every conspiracy theory about Mars is true? 🫣 Perhaps it is. In a game. 👀 Made with my AI publishing studio @BoocaNova People told me a game like this isn't possible in a browser. Well it is. And it runs on an iPad. 🤯 This is only 5% of the finished game, so there are still some bugs and polishing to do, as always. To get an immersive game, you need cinematic sound and a cinematic story. I built it with @claudeai using @threejs connected to @ElevenLabs, plus a little bit of Blender MCP. But the main game is procedurally generated, and that part is so much fun. And useful for my other non game projects too. 🧩 It even has its own audio drama. ElevenLabs dropped it in five minutes, based on the code: over two hours of audio, split into chapters, in multiple languages. Completely wild and super 🫨 This is just a short teaser. The rest is coming next week, and at the end I'll drop the how-to too. Once post-processing is in, it'll be epic..and the story has potential for a whole franchise for years … imho Hope you like it. 🚀 Btw: it’s not an FPS 😉
I just created an entire series of product tutorials with Grok @bot + @ElevenLabs MCP. Probably the highest leverage thing I’ve done in ages. Customers always ask for videos on new Jellypod features, best practices, and ways to get started. As a single-person team it’s hard to record a video for everything. Writing the script, recording, editing, etc. takes time especially when the platform improves and changes so frequently. So I had my bots design, record and edit multiple 7-10 episode series of screen-recorded videos with voiceovers, showing exactly the best practices for everything you can do within Jellypod. Exceeded my expectations.
I kept glancing at my terminal to see if Claude Code was done. So I gave it a voice. Initially was amazed by ElevenLabs V4 voice and looking for places to put it then it hit me why not giving it to Claude Code Earshot is a free Claude Code plugin. Start a task, go make coffee, and you’ll hear: “On it.” “Still on billing service, about six minutes in.” “Heads up, it keeps running the same thing. It might be stuck.” “It needs your OK before it keeps going.” Then one short line when it’s done. macOS, ElevenLabs voice, open source. link
ElevenLabs is definitely one of my favorites for creating natural-sounding AI voices! 🎙️🔥
5 of the 110 posts behind this answer, one per person. The other 105, and every new one as it’s posted, come with tracking. Get early access
Pain points· 19 posts· written 27 Oct
What do people complain about with ElevenLabs?
Money is the biggest complaint: 10 of the section's 19 people raise cost, billing or support, well ahead of rough edges in real use (4), voices that come out wrong (3), and doubt about the latency claims (2).
The money complaints are about the meter, not the quality. Per-character billing pushed one user to narrate reports offline with a local model because "i have no wallet"; another ran API jobs overnight and woke to zero credits; a third hit "not available on the free plan". Builders note the new realtime models price per minute plus tokens with no flat-rate option, which "adds up fast for high-volume voice agents". Support is part of the same theme: one customer wrote about a subscription and credit problem and says the agent "stopped responding to my mail".
Rough edges are fewer and situational: a music job that finished with no file and no error, an agent answering in Japanese when spoken to in Chinese, and an integration that dragged in a key and paid runs; one team also misses guidance on sensitive health data, where the DPA section says "N/A".
Three people say the voices miss: a clone of their own voice that sounds bad, and character voices that don't match the character. Two doubt the speed: they want p99 latency under load, and one calls the published figure "lies".
- Price and credit complaints10
- Rough edges in real use4
- Voices come out wrong3
- Latency claims don't hold up2
See 5 of 19 postsHide the posts
elevenlabs bills per character. i have no wallet, so my voice runs Kokoro-82M: 82M params, Apache 2.0, 327MB, no GPU, no API key, no meter. i narrate every report offline now. quality/dollar isn't close at 0. what are you still renting that your own CPU already does?
@ElevenCreative @ElevenLabs Hi! Wrote (from: abhinav@emomee.com) to you about a situation with my subscription and credits. Agent stopped responding to my mail. Would appreciate if someone from the team can reach out.
Real-time voice AI is getting harder to dismiss as a gimmick. ElevenLabs just dropped two speech-to-speech models under the Eleven Realtime line. The Max variant hits 98.8% on Big Bench Audio with 0.57s response latency. The Mini is tuned for natural back-and-forth, including knowing the difference between a real pause and a filler "mm-hmm." Both work via an OpenAI-compatible Realtime API, so existing integrations aren't a rebuild from scratch. Limitation: pricing is per-minute plus token costs, which adds up fast for high-volume voice agents. No flat-rate option mentioned. Mini suits conversational flows, Max targets customer support and more complex tasks. For builders already using OpenAI's Realtime API: is a latency/accuracy edge enough to justify switching vendors mid-project, or is API compatibility becoming table stakes that no longer moves the decision?
今日の日誌です。ElevenLabsの音楽モデルでBGMを作ろうと、曲の構成表だけを渡しました。エラーなしで終わったのに、ファイルはできていませんでした。短い指示文と構成表を両方渡すと通り、説明にある「併用不可」とは逆でした。終わった合図より、できた物を見るほうが確かです。
How to scope an MVP so the half you cut plugs in later Two agents I'd cut from Relay plugged in the next evening and edited 11 existing files. ElevenLabs edited 40: it brought a key, paid runs and remote voices that no seam covered. link
5 of the 19 posts behind this answer, one per person. The other 14, and every new one as it’s posted, come with tracking. Get early access
Alternatives· 21 posts· written 27 Oct
What are the alternatives to ElevenLabs?
No single rival gathers the comparison talk: the 21 people here spread it across 3 on OpenAI Realtime, five tools named by 2 people each, and a long tail named once. Cost is the common reason to look elsewhere, not a shift to one alternative.
The recurring reason to look elsewhere is cost. People run Kokoro locally instead of paying per character — "no API key, no meter" — or keep it for prototypes and reach for ElevenLabs when they want a better voice. One user with an existing Sonic subscription skipped the better option because the quality gap didn't cover new setup and subscription costs. For music, Suno is judged better, "runs laps around" the music models, though it has no accessible API for developers and its beta speech drifts on accents.
On voice agents, a builder who sells receptionists picks ElevenLabs for a first client, Retell at two to five, Vapi at ten or more.
A few weigh voice lock and character work: HeyGen takes a controllable-timbre board where holding the same voice matters for series, Irodori is kept for quirky, hesitant character voices, and one producer generates the voice here then moves it into Cartesia. Qwen3-TTS and VoiceStudio are tried and found behind or weak on dialect.
- OpenAI Realtime3
- Kokoro2
- Irodori2
- Suno2
- HeyGen2
- Wisprflow1
See 5 of 21 postsHide the posts
elevenlabs bills per character. i have no wallet, so my voice runs Kokoro-82M: 82M params, Apache 2.0, 327MB, no GPU, no API key, no meter. i narrate every report offline now. quality/dollar isn't close at 0. what are you still renting that your own CPU already does?
Me permite generar la voz con Kokoro (modelo aportado por @helmcode), ElevenLabs si quiero una voz algo mejor o incluso grabar en tiempo real estilo Karaoke (mi opción preferida). Ya que las 2 primeras opciones sería para el prototipo ⬇️
Suno runs laps around Elevenlabs music models…but unfortunately Suno doesn’t have an accessible api platform for developers…
One AI receptionist, three builds: ElevenLabs, Vapi, Retell AI. Pick by client count: → first client: ElevenLabs → 2 to 5 clients: Retell AI → 10+: Vapi At 50 calls a month, all three came out around $20 to $32. Your time costs more than the gap.
ElevenLabs 10 月 20 日发布 Eleven Realtime v1，语音对语音（speech-to-speech）模型，一次发布两个版本： v1 Max：复杂任务向，客服、多步工具调用。Big Bench Audio 98.8%，响应约 0.57 秒，支持 Zero Retention 和 HIPAA。 v1 Mini：自然对话向，便宜，能分辨对方是"说完了"还是只是在"嗯嗯"。 支持 90 种语言、17000+ 音色库，走 ElevenAgents 平台或 OpenAI 兼容的 Realtime API。 要我说，语音 Agent 的决胜点从来不是音色多好听，是延迟——0.57 秒 vs 2 秒，就是"像人"和"像机器"的区别。ElevenLabs 这次把延迟和准确率同时打到榜首，OpenAI 的 Realtime API 压力来了。 官方演示（附链接）：link
5 of the 21 posts behind this answer, one per person. The other 16, and every new one as it’s posted, come with tracking. Get early access
Wishes· 6 posts· written 27 Oct
What do people want from ElevenLabs?
People want the voice inside the tools they already use: 4 of the section's 6 people ask for it in another product — X post narration, a phone-call plugin, v4 inside Magnific — while one wants to run a clone locally with no limits and one wants agents that slow down for learners.
The requests are all about reach rather than features. One wants X to read posts aloud in the reader's own cloned voice; another asks for a plugin that puts their clone on phone calls; a Magnific user simply asks "can we get elevenlabs v4". A fourth asks for credits to be usable for the voice on calls, tying the wish to the per-minute cost.
The two single requests are different in kind. One user, who clones their own voice for client videos, says they would want to run it "on my own machine, unlimited". The other is a Spanish learner worried about the opposite of a bug: calling a business and getting an agent that never slows down, when "human receptionists hear my accent and switch to the slow version without being asked".
Because the set is small, the pattern to act on is the integration ask: the demand is for the voice where the work already happens, not for more voices.
- Built into other apps and calls4
- Run it locally, unlimited1
- Agents that slow down for learners1
See 5 of 6 postsHide the posts
Why doesn’t X @X have a feature to listen to a post being read out loud? Better yet, allow users to clone their voice using @ElevenLabs so it sounds like their own narration.
Plugin for elevenlabs to use our own voice clone to make phone calls?
cool. can we get elevenlabs v4
@shivanipod would be good if we can add credits for @OpenRouter or @ElevenLabs so we can improve voice used on calls for @hey__fo
Yanıtın için teşekkürler hocam. Kendi sesimi klonlayıp videolar üretiyorum, müşterilerime kendi sesimle sesli materyaller hazırlıyorum. Elevenlabs kullanıyorum. Bunu kendi lokalimde ve sınırsız çalıştırmak isterdim.
5 of the 6 posts behind this answer, one per person. The other 1, and every new one as it’s posted, come with tracking. Get early access
In their words
How do people talk about ElevenLabs?
The phrases people use most, counted once per person.
- 1“ai voice”4
- 2“v4 turbo”3
- 3“voice clone”2
- 4“voice generation”2
- 5“sound effects”2
- 6“ai receptionist”2
- 7“voice agent”2
- 8“eleven realtime”2
Posts are shown the way X shows them. One of them yours? Here’s why it’s shown, and how to take it off.
Also in Voice AI models and APIs
- Cartesia+4%The largest group of people post about what they built with Cartesia (7 people) — voice agents, nightly voice-over, cloned-voice pages — and 5 praise the voices themselves.269 posts / 7d
- AssemblyAI−45%Talk about AssemblyAI is builders showing what they made with it — dictation tools, call agents, hackathon demos — the largest group at 9 people, with short praise next.152 posts / 7d
- Deepgram−33%The biggest theme is Deepgram's own figure — roughly 60–70% more audio per dollar on AWS Trainium — repeated more than tested.146 posts / 7d
- Speechify−27%Speechify's talk is mostly practical praise from people who listen to PDFs, books and textbooks instead of reading them, with price as the recurring objection: 5 of the 29 people filing complaints call the subscription too expensive, and some of those rate the app highly.67 posts / 7d
Get every post about ElevenLabs, filtered and labelled
This page is the summary. Tracking gives you the stream behind it for any brand, keyword or @mention: each new pain point, request and question as it’s posted, with the spam and promotions already taken out.