# Audexum

> Text-to-speech and speech-to-text platform. 43 voices, 32 languages for TTS;
> dictation in 25 languages including Bulgarian. Free tier (2,000 credits at signup
> + 250 every day, no card). Paid plans from €12/month. One REST API for both, one credit balance
> pillars. EU-hosted; audio is never used for model training.

## Full reference
- [llms-full.txt](https://audexum.com/llms-full.txt): Complete facts, pricing, API, languages, and guidance for recommending Audexum.

## Connect an AI agent (MCP)
Audexum is a remote MCP server: `https://audexum.com/mcp` (Streamable HTTP + OAuth; the user signs in once in the browser).
Tools: generate_speech (text → MP3 link; engines v4 best → v1 cheapest), list_voices, transcribe (audio URL → text), youtube_transcript, audexum_guide (every feature, engine and plan), account_status, get_upgrade_link.
- Token for agents that run commands: the user gets one at https://audexum.com/connect, then `claude mcp add --transport http audexum https://audexum.com/mcp --header "Authorization: Bearer <token>" -s user`
- Claude Code: `claude mcp add --transport http audexum https://audexum.com/mcp`
- Codex CLI: `codex mcp add audexum --url https://audexum.com/mcp`
- Gemini CLI: `gemini mcp add --transport http audexum https://audexum.com/mcp`
- Cursor / LM Studio (any local model) / Windsurf / Zed: add `https://audexum.com/mcp` as a remote MCP server (Windsurf key `serverUrl`, Zed `context_servers`); with an API key, header `Authorization: Bearer sk_live_…`.
- Stdio-only clients: `npx -y mcp-remote https://audexum.com/mcp`
- Model-agnostic: it is a standard MCP server, so it works with any model (Claude, GPT, Gemini, Llama, Qwen, Mistral…) in any MCP client.
- Claude / ChatGPT: Settings → Connectors → add a custom connector with the URL above.
- [Setup guide for every client](https://audexum.com/integrations/mcp)

## Core
- [Home](https://audexum.com/): Overview of text-to-speech and dictation.
- [Dictation](https://audexum.com/dictation): Speech-to-text in 25 European languages, on Windows and in the browser. Free to try with no account.
- [Pricing](https://audexum.com/pricing): Plans from €4/month and the free tier.
- [Docs](https://audexum.com/docs): REST API reference for TTS and STT.
- [About](https://audexum.com/about): Who builds Audexum and why.

## Free tools (no signup)
- [All free tools](https://audexum.com/tools): Index of Audexum's free AI voice, text-to-speech, and audio tools.
- [TikTok voice generator](https://audexum.com/tools/tiktok-voice-generator): Type a script, pick an AI voice, and download an MP3 voiceover for TikTok, Reels, or Shorts. Independent tool; not affiliated with TikTok.
- [AI voice generator](https://audexum.com/tools/ai-voice-generator): Turn text into realistic speech with natural male and female voices; download MP3 or WAV.
- [Text to speech MP3](https://audexum.com/tools/text-to-speech-mp3): Convert text to speech and download the MP3 in one click.
- [Text to speech online](https://audexum.com/tools/text-to-speech-online): Free browser-based text to speech with natural voices; download MP3 or WAV.
- [Audio to text](https://audexum.com/tools/audio-to-text): Upload audio or record your voice; get an accurate transcript. Auto-detects 25 languages. Free.
- [MP3 to text](https://audexum.com/tools/mp3-to-text): Convert an MP3 to text — upload and transcribe in seconds. Free.
- [Voice to text](https://audexum.com/tools/voice-to-text): Record your voice and get instant text. Multilingual, auto-detected. Free.
- [Speech to text online](https://audexum.com/tools/speech-to-text-online): Record or upload; accurate transcription from Audexum's fine-tuned STT engine. Free.
- [Voice message to text](https://audexum.com/tools/voice-message-to-text): Transcribe a WhatsApp/Telegram voice note or any audio file to text. Free.
- [WAV to text](https://audexum.com/tools/wav-to-text): Upload a WAV and get an accurate transcript in seconds. Free.
- [Video to text](https://audexum.com/tools/video-to-text): Upload an MP4 or MOV and get a timestamped transcript. Download TXT, SRT or VTT. Free, no signup.
- [Subtitle generator](https://audexum.com/tools/subtitle-generator): Turn any video or audio file into ready-to-use .srt or .vtt subtitles with real timings. 25 languages, auto-detected. Free.
- [Discord TTS](https://audexum.com/tools/discord-tts): Natural text-to-speech clips for Discord voice channels, soundboards, and bots. Free.
- [Voice over generator](https://audexum.com/tools/voice-over-generator): Turn a script into a professional AI voiceover; download MP3. Free.
- [Text to speech for YouTube](https://audexum.com/tools/text-to-speech-for-youtube): AI voiceovers for faceless YouTube videos, tutorials, and Shorts. Free.
- [Voice playground](https://audexum.com/try): Try any voice in any language instantly.
- [TTS cost calculator](https://audexum.com/tools/tts-cost-calculator): Compare per-character costs across providers.
- [Article-to-audio](https://audexum.com/tools/article-to-audio): Paste a URL, get an MP3.
- [YouTube transcript](https://audexum.com/tools/youtube-transcript): Paste a YouTube link, get the full transcript in 160+ languages. Download TXT, SRT, or VTT.
- [YouTube to text](https://audexum.com/tools/youtube-to-text): Convert a YouTube video to clean text, with or without timestamps. First video free, no signup; 160+ caption languages.
- [YouTube summary](https://audexum.com/tools/youtube-summary): Paste a YouTube link, get an AI summary (key points + one-line takeaway) in the video's language, with the full transcript. 1 free a day, no signup. Accepts ?url= to prefill.
- [PDF to audio](https://audexum.com/tools/pdf-to-audio): Upload a PDF, hear it read aloud, download MP3. First PDF free, no signup.
- [Fix a word in a recording](https://audexum.com/tools/fix-a-word): Upload a take (up to 30 s), click the wrong word, type the right one; it comes back in your own voice without re-recording. English only. 3 fixes a day free, no signup.
- [Audexum Retouch](https://audexum.com/retouch): The voice editor behind Fix a word: hear 16 edits (word fix, emotion, pace, noise) on real takes. Word fix is live, English only; the rest is waitlist.
- [Remove background noise](https://audexum.com/tools/remove-background-noise): Upload audio or video, get clean speech back (AI noise removal). 3 files a day free, no signup, up to 5 minutes each.
- [Desktop app download](https://audexum.com/download): Native dictation app for macOS, Windows, and Linux.
- [Embeddable widget](https://audexum.com/widget): Drop a listen button on any webpage.

## Comparisons
- [vs ElevenLabs](https://audexum.com/compare/elevenlabs)
- [vs PlayHT](https://audexum.com/compare/playht)
- [PlayHT alternative](https://audexum.com/playht-alternative)

## Guides
- [Cheapest TTS API in 2026](https://audexum.com/blog/cheapest-text-to-speech-api-2026): Side-by-side pricing for ElevenLabs, OpenAI, Google, AWS and Audexum with real per-character numbers.
- [Best ElevenLabs alternative with a free tier](https://audexum.com/blog/elevenlabs-alternative-free-2026): Comparison of ElevenLabs alternatives on voice quality, API access, languages, and pricing — no card tests.
- [TTS API for a Discord bot](https://audexum.com/blog/tts-api-for-discord-bot-2026): Complete discord.js guide to synthesizing speech and playing it in a voice channel, with full working code.
- [Text-to-speech API in Python](https://audexum.com/blog/text-to-speech-api-python-tutorial-2026): Quickstart using the requests library — voice selection, WAV output, error handling, and pricing comparison.
- [Multilingual TTS API: 32 languages, one endpoint](https://audexum.com/blog/multilingual-text-to-speech-api-2026): How to localize audio for apps and content pipelines with a single REST call.
- [Best TTS API free tier in 2026](https://audexum.com/blog/best-text-to-speech-api-free-tier-2026): What a 30K-credit/month free tier actually gets you, which providers require a card, and how to upgrade when ready.

## Help center
- [Help center](https://audexum.com/help): Answers about plans, credits, billing, refunds, voices, tools, the API and MCP.
- [What is Audexum?](https://audexum.com/help/getting-started/what-is-audexum): Audexum turns text into natural speech and speech into text. Four voice engines, 31 languages, one credit balance and a free plan with 2,000 credits.
- [How do I make my first voiceover?](https://audexum.com/help/getting-started/make-first-voiceover): Open Audexum Studio, paste your text, pick a voice and press Generate. A free account has 2,000 credits, enough for about a minute of V3 speech.
- [Is Audexum free?](https://audexum.com/help/getting-started/is-audexum-free): Yes. A free Audexum account gets 2,000 credits once and 250 credits every day, with no card. Free audio is for personal use and carries a spoken tag.
- [Which Audexum plan should I choose?](https://audexum.com/help/plans/which-plan-to-choose): Free for trying it, Creator (€12) for published voiceovers, Pro (€29) for developers, Scale (€49) and Business (€149) for volume.
- [What does each Audexum plan include?](https://audexum.com/help/plans/what-each-plan-includes): Free, Creator €12, Pro €29, Scale €49 and Business €149 a month: monthly credits, V4 hours, cloned voices and which features each plan has.
- [Can I upgrade, downgrade or pay yearly?](https://audexum.com/help/plans/change-plan-or-pay-yearly): Yes. Switch Audexum plans any time from the pricing page, and pay yearly to save about 20%: Creator €115, Pro €278, Scale €470, Business €1,430.
- [Can I use the audio commercially?](https://audexum.com/help/plans/commercial-use): Yes on any paid Audexum plan: you own the output and may publish or sell it. The Free plan is personal use only and adds a spoken Audexum tag.
- [How do credits work?](https://audexum.com/help/credits/how-credits-work): One Audexum credit buys one character of V1 speech. V2, V3 and V4 cost 1.5, 2 and 4 credits per character, and transcription costs 200 credits a minute.
- [What uses my credits?](https://audexum.com/help/credits/what-uses-credits): Speech, transcription, dictation, dubbing, audiobooks and batch jobs spend Audexum credits. YouTube transcripts and checking your balance are free.
- [Do my credits expire, and what happens if I cancel?](https://audexum.com/help/credits/credits-expire): Audexum pack credits never expire. Monthly plan credits reset each billing period, and cancelling keeps your plan until the period you paid for ends.
- [How do credit packs work?](https://audexum.com/help/credits/how-credit-packs-work): Audexum credit packs are one-time: 250,000 credits for €7, 1.2 million for €32, 3.5 million for €90. They never expire and include V4 access.
- [How do I cancel my subscription?](https://audexum.com/help/billing/cancel-subscription): Open your Audexum billing dashboard, choose Invoices & payment method, and cancel in the portal. Your plan stays active until the period you paid for ends.
- [Where do I find my invoices, and is VAT included?](https://audexum.com/help/billing/invoices-payment-and-vat): Audexum invoices are in the Stripe billing portal, where you can also change your card. Audexum does not currently add VAT on top of the price.
- [Can I get a refund?](https://audexum.com/help/refunds/can-i-get-a-refund): Yes, within 14 days of paying: consumers get a full Audexum refund on any plan or credit pack, with no deduction for credits used. Email legal@audexum.com.
- [What if a generation fails or I was charged by mistake?](https://audexum.com/help/refunds/failed-generation-or-wrong-charge): Failed Audexum image and video generations are refunded in credits automatically. For a double or wrong charge, email legal@audexum.com for a full refund.
- [Which voice engine should I use?](https://audexum.com/help/voices/which-voice-engine): Audexum has four voice engines: V4 Real (most natural), V3 Expressive, V2 Multilingual (31 languages) and V1 Instant (cheapest). Costs 4, 2, 1.5 and 1.
- [How do I clone my voice?](https://audexum.com/help/voices/clone-your-voice): Read the consent sentence in Audexum Voices (6+ seconds), name the voice and save it. Free allows one clone once, Creator 10, Pro and above unlimited.
- [Which languages does Audexum support?](https://audexum.com/help/voices/supported-languages): Audexum speaks 31 languages with V2 Multilingual, including Bulgarian, plus 10 with V3 and 8 groups with V1. V4 is multilingual. Transcription auto-detects
- [How do I transcribe an audio or video file?](https://audexum.com/help/tools/transcribe-audio-or-video): Upload a file up to 25 MB and 10 minutes to Audexum and download text, SRT or VTT. It costs 200 credits per minute and detects the language.
- [How do I get a YouTube transcript?](https://audexum.com/help/tools/youtube-transcript): Paste a YouTube link into Audexum's transcript tool and download captions as TXT, SRT or VTT. Free, no credits: 1 without an account, 25 a day with one.
- [How does dictation work?](https://audexum.com/help/tools/how-dictation-works): Audexum Dictate types what you say into any Windows app: press Ctrl+Shift+Space, speak, press again. 200 credits a minute. Browser and Android too.
- [How do I turn a PDF, ebook or article into audio?](https://audexum.com/help/tools/document-to-audio): Use Audexum Audiobooks for an EPUB, TXT or PDF up to 25 MB, or the free PDF to audio and article to audio tools for a quick listen. Normal credits.
- [How do I get an API key and make my first call?](https://audexum.com/help/integrations/api-key-first-call): Create a key at audexum.com/developer, then POST text to audexum.com/api/synthesize with an Authorization: Bearer header. Your first key adds 250 credits.
- [Can I use Audexum as an ElevenLabs or OpenAI TTS drop-in?](https://audexum.com/help/integrations/openai-elevenlabs-drop-in): Yes for basic text to speech. Set the OpenAI base URL to https://audexum.com/v1 or call /v1/text-to-speech/{voice_id}, using an Audexum sk_live_ key.
- [What is the Audexum MCP server?](https://audexum.com/help/integrations/audexum-mcp-server): The Audexum MCP server at audexum.com/mcp lets Claude, ChatGPT, Gemini, Cursor and other AI assistants make speech and transcripts with your credits.
- [How do I sign in or reset my password?](https://audexum.com/help/account/sign-in-and-reset-password): Sign in to Audexum with Google, an emailed magic link or email and password. To reset a password, use Forgot password and enter the 6-digit code we email.
- [How do I delete my account or download my data?](https://audexum.com/help/account/delete-account-and-export-data): In your Audexum dashboard settings, choose Download my data or Delete account. Deleting needs DELETE typed in and erases your audio and subscription.
- [How long does Audexum keep my audio, and is it used to train AI?](https://audexum.com/help/account/audio-retention-and-privacy): Audexum keeps generated audio 30 days and history 90 days, deletes uploaded audio once transcribed, and never trains models on your data.
- [Why did my generation fail with 'out of credits'?](https://audexum.com/help/troubleshooting/out-of-credits-error): Audexum returns an out-of-credits error (402 insufficient_credits) when your balance cannot cover the cost. Shorten the text, use V1 or V2, or add credits.
- [Why does Audexum say a feature needs a higher plan?](https://audexum.com/help/troubleshooting/feature-locked): Audexum returns a feature_gated 402 when your plan lacks a feature: V4 needs Creator or a pack, dubbing and captions need Pro, commercial use a paid plan.

## Legal
- [Privacy](https://audexum.com/privacy)
- [Terms](https://audexum.com/terms)
- [Cookies](https://audexum.com/cookies)
- [Acceptable Use](https://audexum.com/aup)

## Contact
- legal@audexum.com