[
    {
        "role": "user",
        "content": "What%20is%20the%20most%20costly%20Elevelenlabs%20plan%20for%20home%20users%20and%20as%20I%20created%20an%20agent%20based%20on%20an%20audio%20drama%20I%20created%20hhow%20long%20can%20I%20talk%20to%20it%20for%20and%20how%20do%20I%20update%20it%27s%20information%20like%20the%20brain%20of%20it%20like%20the%20audio%20drama%20transcripts%3F"
    },
    {
        "role": "assistant",
        "content": "Here\u2019s a breakdown of the **most expensive home/user-level plan** on ElevenLabs\u2014and how long you can talk to your custom agent (based on an audio drama you've made), plus how to update its \u201cbrain\u201d (knowledge base) like transcripts, scripts, etc.\n\n---\n\n## \ud83d\udcb8 Most Costly Plan for Home Users\n\nIf by \"home users\" you mean individual creators (not businesses or enterprises), the top-tier plan that makes sense for a single creator is the **Business** plan. Here\u2019s how the plans stack up, and what Business offers:\n\n| Plan | Monthly Price | Included Credits* | Key Features Over Lower Tiers |\n|---|---|---|---|\n| Free | $0 | 10,000 credits | Basic TTS/STT, non-commercial license, most features limited ([elevenlabs.io](https://elevenlabs.io/pricing?utm_source=openai)) |\n| Starter | ~$6/mo | 30,000 credits | Instant voice cloning, commercial license, more projects ([elevenlabs.io](https://elevenlabs.io/pricing?utm_source=openai)) |\n| Creator | ~$22/mo | **121,000** credits | Professional voice cloning, more quality & privileges ([elevenlabs.io](https://elevenlabs.io/pricing?utm_source=openai)) |\n| Pro | ~$99/mo | 600,000 credits | High audio quality, larger output options ([elevenlabs.io](https://elevenlabs.io/pricing?utm_source=openai)) |\n| Scale | ~$299/mo | 1,800,000 credits | Multi-seat workspaces, team collaboration ([elevenlabs.io](https://elevenlabs.io/pricing?utm_source=openai)) |\n| Business | ~$990/mo | 6,000,000 credits | Maximum \u201chome/creator level\u201d capacity, many pro features ([elevenlabs.io](https://elevenlabs.io/pricing?utm_source=openai)) |\n\n\\*Credits are used across all ElevenLabs products (TTS, STT, agents, etc.) ([elevenlabs.io](https://elevenlabs.io/pricing?utm_source=openai))\n\nSo in short: **Business ($990/mo, 6M credits)** is the \u201ctop\u201d home/creator plan before you get into fully custom Enterprise-level contracts.\n\n---\n\n## \ud83d\udde3\ufe0f How Long Can You \u201cTalk\u201d to Your Agent?\n\nDepends on two things:\n\n1. **What you mean by \u201ctalk\u201d** \u2013 audio conversation? text conversation? continuous hours?\n2. **How many credits/budget you have** (depends on your plan).\n\nElevenAgents agents let you do **voice-based conversations**. The conversation length is primarily limited by:\n\n- Your plan\u2019s credit balance (voice data + speech-to-text costs).\n- System limits like maximum lengths per request.\n- Likely practical limits: performance, quality lag, etc. ([elevenlabs.io](https://elevenlabs.io/docs/help-center/product/eleven-agents/what-can-i-create-with-eleven-agents?utm_source=openai))\n\nElevenAgents **default models** have built-in knowledge (pretrained). To add your audio drama transcripts so the agent knows these specifics, you use a **knowledge base**.\n\nSo: you *can* talk until credits run out, or until a conversation hits the agent\u2019s limit (e.g. maximum per-turn context). There\u2019s no fixed \u201ctalk time\u201d \u2014 it\u2019s driven by usage.\n\n---\n\n## \ud83e\udde0 Updating Your Agent\u2019s Brain (Knowledge Base)\n\nHere\u2019s how you give your agent the transcripts, scripts, or any domain-specific info (like your audio drama), so it \u201cknows\u201d and uses them:\n\n1. **Create a Knowledge Base** in your ElevenAgents dashboard. ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/knowledge-base?utm_source=openai))  \n   This is where you store documents (text, transcripts, HTML, etc.) that the agent can refer to in conversation.\n\n2. **Add Documents** \u2013 you can upload files (TXT, PDF, DOCX, Markdown, etc.), paste text manually, or pull in content from URLs. ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/knowledge-base/manage-documents?utm_source=openai))\n\n3. **Attach Knowledge Base to Agent** \u2013 once documents are in your KB, you link them to your agent so they are part of its context. ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/knowledge-base/manage-documents?utm_source=openai))\n\n4. **Editing / Refreshing** \u2013 if you update a transcript (say you've changed the script or added new episodes), you can:\n\n   - **Edit the document directly** (inline editing) or replace the file you uploaded. ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/knowledge-base/manage-documents?utm_source=openai))  \n   - If the document came from a URL and you enabled auto-sync, it'll fetch the updated content automatically. Or manually refresh. ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/knowledge-base/manage-documents?utm_source=openai))  \n   - The system re-indexes your document, recalculating embeddings / RAG data so the agent's responses use the new content. ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/knowledge-base/manage-documents?utm_source=openai))\n\n5. **Usage Modes** \u2013 small docs get put \u201cfully in context\u201d directly in the prompt. Big ones use RAG (Retrieval Augmented Generation), where only relevant parts are pulled from KB during a conversation. ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/knowledge-base?utm_source=openai))\n\n---\n\n## \u2705 TL;DR\n\n- Most costly creator/home plan is **Business** (~$990/month, 6 million credits). Above that it\u2019s likely enterprise. ([elevenlabs.io](https://elevenlabs.io/pricing?utm_source=openai))  \n- How long you can \u201ctalk\u201d depends on credits, conversation limits, and usage\u2014no fixed cap beyond these.  \n- To update the agent\u2019s \u201cbrain\u201d:\n\n  1. Build a knowledge base.\n  2. Upload or edit your transcripts/scripts.\n  3. Attach to the agent.\n  4. Refresh or auto-sync if needed.\n\n---\n\nIf you want, I can compute **how many hours** your credits would give you for voice conversations, given your budget or how many episodes you have. Do you want me to crunch that for you?"
    },
    {
        "role": "user",
        "content": "I%20am%20in%20the%20%246%20per%20month%20and%20get%2015%20minutes%20then%20it%20cuts%20off%20so%20with%20the%20creater%20plan%20what%20would%20my%20limits%20be%20and%20does%20the%20service%20have%20a%20memory%20as%20I%20have%20to%20remind%20the%20agent%20when%20I%20talk%20to%20it%20using%20voice%20as%20I%20am%20totally%20blind%20and%20would%20like%20to%20have%20it%20remember%20things%20like%20past%20conversations%3F"
    },
    {
        "role": "assistant",
        "content": "Upgrading from the $6/month (\u201cStarter\u201d) plan to the **Creator** plan gets you more capacity\u2014but there are still some limits. Also, ElevenLabs offers features to help your agent *remember* past conversations (so you don\u2019t have to re-repeat things every time you talk). Below are the details.\n\n---\n\n## \u2699\ufe0f Creator Plan: What You Get\n\nIf you move up to the **Creator** plan (currently $22/month), you\u2019ll have:\n\n- **121,000 credits/month** to spend across all voice, speech-to-text, etc. ([elevenlabs.io](https://elevenlabs.io/pricing?utm_source=openai))  \n- More voice cloning options \u2014 including *Professional Voice Cloning*. ([elevenlabs.io](https://elevenlabs.io/pricing?utm_source=openai))  \n\n**What this means for talk-time:**  \nBecause credit cost depends on the model (voice, speech-to-text, etc.), the actual minutes you get depends on what features you use. As a rough estimate:\n\n| Usage | Approximate Credit Cost / Time |\n|---|---|\n| Using a \u201cstandard\u201d voice + normal speech-to-text | You might get ~100-130 minutes of voice conversation a month with Creator, depending on how chatty the agent is. ([toolincomelab.com](https://toolincomelab.com/articles/ai-voice-generator-cost-per-hour/?utm_source=openai)) |\n| If you use higher-quality voices, heavier speech processing, voice cloning, etc. | Costs will be higher per minute, so you\u2019ll get fewer minutes. |\n\nAlso, Creator and higher plans allow **Usage-Based Billing** (paying for extra credits when you use them over your monthly allotment) so you aren\u2019t locked out if you go over. ([help.elevenlabs.io](https://help.elevenlabs.io/hc/en-us/articles/24351324241937-What-is-the-price-for-usage-based-billing?utm_source=openai))\n\n---\n\n## \ud83e\udde0 Memory & Retention: Will Your Agent \u201cRemember\u201d You?\n\nYes \u2014 ElevenLabs offers features so your agent can retain memory of past conversations and use them in future chats. Here\u2019s how that works:\n\n### Conversation History & Transcripts\n\n- ElevenLabs stores transcripts and audio recordings of conversations by default. You can search through what\u2019s been said previously. ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/operate/overview?utm_source=openai))  \n- **Retention Settings** let you control how long that data is kept (e.g., 30 days, 90 days, 365 days, or indefinitely). ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/privacy/retention?utm_source=openai))\n\nSo if you tell the agent something via voice (like \u201cI\u2019m blind, remind me\u2026\u201d), future conversations can reference that if the retention is active.\n\n### Knowledge Base (Your \u201cBrain\u201d Outside Transcripts)\n\nTo add more structured, reliable memory (like your audio drama transcripts, new content, etc.):\n\n- You can upload documents (transcripts, text, etc.) to a **Knowledge Base** for each agent. ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/knowledge-base?utm_source=openai))  \n- Two main styles of using KB documents:\n\n  1. **Full Context** \u2013 small documents can be put entirely into the context (prompt) each turn.  \n  2. **RAG (Retrieval-Augmented Generation)** \u2013 for larger documents, the system picks only the relevant parts to pull into response building. ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/knowledge-base?utm_source=openai))\n\n- With the Creator plan you are limited in how much total knowledge base size you can index. Non-enterprise users can put up to **20 MB** of documents into the knowledge base. ([elevenlabs.io](https://elevenlabs.io/docs/help-center/product/eleven-agents/what-can-i-create-with-eleven-agents?utm_source=openai))\n\n---\n\n## \u2714\ufe0f What to Do So You Don\u2019t Have to Remind Every Time\n\nTo make conversations smoother (especially since you\u2019re using voice and are blind), you can set up your agent for better memory:\n\n1. **Enable retention** in the agent settings. Set it long enough so your past talks are still stored.  \n2. **Tell the agent important facts** (your preferences, ongoing story context) and upload transcripts/scripts to the knowledge base. That way the agent can pull up these parts when needed.  \n3. If possible, use RAG for the transcripts, so even long scripts are used efficiently.  \n4. Use system prompts \u2014 these are instructions you include when creating or editing the agent, telling it \u201cremember that the user is blind,\u201d etc. That helps initial setup. ([elevenlabs.io](https://elevenlabs.io/docs/help-center/product/eleven-agents/what-can-i-create-with-eleven-agents?utm_source=openai))  \n\n---\n\nIf you like, I can check *exactly how many hours* of voice conversation your remaining Creator credits would give *you* under different settings (standard voice vs high quality). Would that be helpful?"
    },
    {
        "role": "user",
        "content": "I%20used%20instant%20voice%20cloning%20so%20how%20long%20do%20I%20get%20with%20talk%20time%20as%2015%20minutes%20on%20%246%20is%20rather%20small%3F"
    },
    {
        "role": "assistant",
        "content": "Here\u2019s a clearer estimate of how much talk time you\u2019d likely get with the **Creator plan** (about $22/month) using Instant Voice Cloning on ElevenLabs, along with factors that affect that number.\n\n---\n\n## \ud83d\udd0d What the Creator Plan Gives You\n\n- **Credits per month**: 121,000 credits. ([elevenlabs.io](https://elevenlabs.io/pricing?utm_source=openai))  \n- Other features: Professional Voice Cloning slot, higher quality voices, more flexibility. ([elevenlabs.io](https://elevenlabs.io/voice-cloning?utm_source=openai))\n\n---\n\n## \u23f1\ufe0f How Many Minutes You Might Get\n\nCredit usage depends on what you\u2019re doing. Instant Voice Cloning generation via text-to-speech (TTS) costs roughly **1 credit per character**. To estimate minutes from credits, it helps to know how many characters are in one minute of natural speech.\n\n- A typical voice-over minute has somewhere around **800\u20131,000 characters** when using a standard voice model. ([toolchase.com](https://toolchase.com/blog/elevenlabs-pricing-2026/?utm_source=openai))  \n- In recent tests, with the *Creator* plan and \u201cnormal\u201d voices, people saw about **121\u2013147 finished minutes of narration per month**, assuming each script take is final (no big redos or editing costs) and with the standard TTS/voice model. ([toolproven.com](https://toolproven.com/blog/elevenlabs-credits-explained?utm_source=openai))\n\nSo with 121,000 credits:\n\n- If you use standard Multilingual v2 voice (one credit per character), you might get **~120\u2013130 minutes** of final audio per month if everything goes smoothly (clear script, no re-takes).  \n- If you choose a simpler model like Flash or Turbo (lighter/faster, though less polished), you may stretch that time further\u2014probably **150\u2013250 minutes** in ideal conditions. ([lokeshkapoor.com](https://lokeshkapoor.com/blog/elevenlabs-pricing?utm_source=openai))\n\n---\n\n## \ud83d\udd01 Why Your $6 Plan Feels Short\n\nYou said on the $6/month Starter plan you get only **15 minutes**, which is much less than what people expect. This can be due to:\n\n- Using higher credit-cost features (voice cloning, high-fidelity voices).  \n- Extra costs for speech-to-text (which also spends credits).  \n- Possibly using a voice with a **credit multiplier** (if someone who made a Professional Voice Clone set it higher). ([help.elevenlabs.io](https://help.elevenlabs.io/hc/en-us/articles/26050485079057-What-are-custom-rates-and-credit-multipliers?utm_source=openai))\n\n---\n\n## \u2705 TL;DR Estimate for You\n\nIf you upgrade to **Creator ($22/mo)** and use Instant Voice Cloning with a standard voice, clean speech:\n\n- Most realistic usage: **~120 minutes** per month of finished, final voice conversation or narration.  \n- If using simpler voices or accepting lower fidelity: maybe **150\u2013200 minutes**.\n\n---\n\nIf you want, I can compare how much of your Creator credit goes to different parts of a voice conversation\u2014like voice generation + speech-to-text + context/model switching\u2014so you can better plan how to use it efficiently. Do you want me to do that?"
    },
    {
        "role": "user",
        "content": "It%20is%2015%20minutes%20then%20I%20have%20to%20restart%20the%20conversation%20and%20get%20another%2015%20minutes%20so%20I%20am%20wondering%20why%20this%20is%20and%20with%20creator%20would%20I%20get%20more%3F"
    },
    {
        "role": "assistant",
        "content": "Here\u2019s what I found on why your conversation stops after about 15 minutes on your current plan, and how those limits change with the Creator plan\u2014\n\n---\n\n## \u23f3 Why Your Conversation Cuts Off at 15 Minutes\n\nWhat you\u2019re experiencing (about 15 minutes then being forced to restart) is likely caused by **a conversation duration limit** built into the system. Some relevant details:\n\n- ElevenLabs has a \u201cMaximum conversation duration\u201d setting for agent calls. The *default* is **600 seconds**\u2014that\u2019s 10 minutes. ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/conversation-flow?utm_source=openai))\n- Some channels have separate timeouts for inactivity. For example, on WhatsApp message-based conversations there is a **15-minute inactivity timeout** (if the agent does not receive input or respond). ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/conversation-flow?utm_source=openai))\n\nSo it\u2019s not necessarily your money plan that\u2019s capping you at 15\u2014part of it could be due to default configuration or channel-specific timeout.\n\n---\n\n## \ud83d\udca1 What Changes with the Creator Plan\n\nWhen you upgrade to the **Creator** plan, you get more \u201cagent minutes\u201d included in your subscription (meaning more time you can talk via voice with your agent) and generally less frequent cutoffs\u2014but some system settings might still apply. Here\u2019s what that means:\n\n- For agent-based voice usage, the Creator plan includes **~275 talking minutes/month** (depending on channel and usage) under normal conditions. ([agenticindex.io](https://agenticindex.io/vendors/elevenlabs?utm_source=openai))  \n- The Creator plan also increases the concurrency limits (how many conversations can run simultaneously) which allows more flexibility. ([help.elevenlabs.io](https://help.elevenlabs.io/hc/en-us/articles/31601651829393-How-many-ElevenLabs-Agents-formerly-Conversational-AI-requests-can-I-make-and-can-I-increase-it?utm_source=openai))  \n- However, even on paid plans, you must still respect the \u201cmaximum conversation duration\u201d setting. This means each continuous conversation can\u2019t run longer than the maximum your agent is configured for. You or someone (possibly via dashboard/agent settings) may need to raise the maximum conversation duration from the default if that\u2019s what's causing the cut.\n\n---\n\n## \ud83d\udd27 What You Can Do\n\nTo improve your experience so you don\u2019t have to restart every 15 minutes:\n\n1. **Upgrade to Creator** \u2013 that gives you more included agent minutes (so you won\u2019t hit monthly limit fast), and higher capacity overall.  \n2. **Check and adjust the \u201cMaximum conversation duration\u201d setting** in the Conversation Flow / Advanced settings in your agent\u2019s dashboard. You may be able to raise it higher than the default (for example the setting can be up to 7,200 seconds, which is **2 hours**). ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/conversation-flow?utm_source=openai))  \n3. **Enable call queuing (if needed)** \u2014makes sure new calls are queued instead of rejected when concurrency is maxed. ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/guides/call-queueing?utm_source=openai))  \n4. **Use KB + memory features** so past conversations are remembered, which reduces the need to re-establish context when you do have to restart.  \n\n---\n\nIf you\u2019d like, I can tell you *specifically* how long conversations are allowed for agents in your region / subscription type, and walk you through how to change the \"maximum conversation duration\" so yours can go longer. Do you want me to do that?"
    },
    {
        "role": "user",
        "content": "yes%20I%20am%20in%20the%20US%20and%20want%20the%20longest%20setup%20possible%3F"
    },
    {
        "role": "assistant",
        "content": "Here are the current details for users in the US on ElevenLabs who want to maximize voice conversation length, along with step-by-step instructions for setting the longest possible duration:\n\n---\n\n## \ud83d\udd12 Maximum Conversation Duration\n\n- By default, new agents have a **max conversation duration** of **600 seconds** (10 minutes). ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/conversation-flow?utm_source=openai))\n- You can change that setting. The valid range is **60 seconds up to 7,200 seconds** (that\u2019s **2 hours**) via agent settings. ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/conversation-flow?utm_source=openai))  \n- Note: Certain channels have separate rules. For example, WhatsApp voice calls follow this limit, but text conversations on WhatsApp have a default **15-minute inactivity timeout** which can end a conversation early. Enterprise customers can request adjustments. ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/conversation-flow?utm_source=openai))\n\n---\n\n## \u2699\ufe0f Limits and Upgrades Under the Creator Plan\n\n- Creators can make up to **10 concurrent conversations** at a time. ([help.elevenlabs.io](https://help.elevenlabs.io/hc/en-us/articles/31601651829393-How-many-ElevenLabs-Agents-formerly-Conversational-AI-requests-can-I-make-and-can-I-increase-it?utm_source=openai))  \n- Creator plan also includes large monthly credit allotments that let you pay for longer calls without running out of credits quickly. ([reddit.com](https://www.reddit.com/r/ElevenLabs/comments/1tha6ma/creator_account_limits/?utm_source=openai))  \n- Usage-based billing is available so you can use more than the monthly included credits, though at extra cost. Set thresholds so you don\u2019t get surprise charges. ([help.elevenlabs.io](https://help.elevenlabs.io/hc/en-us/articles/27378406011409-What-is-usage-based-billing?utm_source=openai))\n\n---\n\n## \ud83d\udd27 How to Set Up the Longest Possible Conversation Time\n\nHere\u2019s how to raise the max conversation duration via the ElevenLabs dashboard:\n\n1. Log in to your ElevenLabs account.\n2. Go to the **Agents** section and select the agent you want to extend.\n3. Navigate to **Advanced** \u2192 **Conversation Flow** settings. ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/conversation-flow?utm_source=openai))  \n4. Find *Max conversation duration*. Change the value from the default (600 seconds) to the maximum allowed: **7,200 seconds**. That will allow up to a 2-hour continuous voice conversation. ([elevenlabs.io](https://elevenlabs.io/docs/eleven-agents/customization/conversation-flow?utm_source=openai))  \n5. Save your settings.\n\nAlso, if you want, fill in the **max_conversation_duration_message** field with a message like: \u201cConversation limit reached, please reconnect to continue.\u201d This message will play when the limit is hit. ([releases.sh](https://releases.sh/release/rel_YSKS9lQZiCKtoGx_XFcOB?utm_source=openai))\n\n---\n\nIf you set it to 7,200 seconds, under the Creator plan with enough credits, you should be able to talk continuously for **up to 2 hours** in that single voice conversation\u2014assuming you don\u2019t hit other limits (credits, concurrency, inactivity).\n\nWould you like me to estimate how many credits a 2-hour conversation uses under different speech models (standard voice vs high fidelity) so you can plan ahead?"
    }
]