Beyond Presence AI in 2026: Pricing, Plan Limits and When to Pick It

Beyond Presence sells credits on monthly plans priced in euros. As of 30 October 2026: Free (2,000 credits), Starter at €49 (14,000), Growth at €149 (74,500) and Scale at €349 (200,000), with Enterprise by quote. A speech-to-video minute, the face for your own agent, uses 50 credits, and a Managed Agent minute uses 100. So Starter covers 280 face-only minutes or 140 agent minutes, Growth 1,490 or 745, and Scale 4,000 or 2,000. Included and extra minutes cost about the same: €0.175, €0.10 and €0.0875 a face-only minute on the three paid plans, and €0.35, €0.20 and €0.175 an agent minute. Extra minutes bill only if usage-based billing is on; otherwise new sessions are refused once the allowance runs out. The plans cross over. Starter plus extra minutes passes Growth's €149 at about 851 face-only minutes a month, and Growth passes Scale at about 3,490. Enterprise rates, Beyond Presence says, fall to €0.03 a minute at scale.

With your own voice agent, 1,000 face-only minutes a month cost €149 on Growth, which includes 1,490; Starter plus 720 extra minutes would be €175. At 10,000 minutes the cheapest option is Scale: €349 plus 6,000 extra minutes at €0.0875, or €874. Realistic Avatars costs $20 and $200 for the same minutes, billed per second. With the conversation included, Managed Agents cost €200 for 1,000 minutes (Growth plus 255 minutes at €0.20) and €1,749 for 10,000 (Scale plus 8,000 at €0.175), against $60 and $600 on Voice Agents. Beyond Presence bills in euros and we bill in dollars, so the gap moves a little with the exchange rate; we haven't converted. Past these volumes both sides quote: Beyond Presence Enterprise from €0.03 a minute, and our enterprise and high-volume contracts from $0.005.

Beyond Presence sells two products on the same credits. The Speech-to-Video API takes your agent's spoken audio, usually its TTS output, and returns a 1080p avatar stream. You run speech recognition, the LLM, the voice and turn-taking, and connect through its LiveKit plugin for Python or JavaScript, Pipecat's BeyTransport or your own ElevenLabs agent. That's the product to set against Realistic Avatars, which does the same job over WebRTC or in your LiveKit room. Managed Agents run the whole conversation: a Cartesia voice by default, an OpenAI-compatible LLM of your choice, a document knowledge base, camera vision from Starter, function calling from Scale and a wakeup mode for group calls. Our counterpart is Voice Agents: a prompt of up to 20,000 characters, up to 20 tools that call your URLs and 160 voices in 15 languages, shared as a link or an embed, for $0.06 a minute against Beyond Presence's €0.175–0.35. Camera vision, a document knowledge base and meeting bots are Beyond Presence extras.

Beyond Presence's Free plan gives 40 face-only minutes or 20 agent minutes a month, one session at a time, stock avatars and sessions capped at 3 minutes, and its comparison table marks it as not for commercial use. Paid plans allow 10, 25 or 50 sessions at once (Starter, Growth, Scale), counted across the whole account, and a request over the limit gets a 429 error. In October 2026 speech-to-video sessions went from 3 hours to 24, and they end after an hour with no speech; idle time still holds a concurrency slot. A widget that must take 30 calls at once needs Scale at €349 a month, whatever its minutes. Realistic Avatars has no plans and no limit on sessions at once. New accounts get $1 of credits (50 minutes) with commercial use included, and each session can run 3 hours.

Beyond Presence's custom avatars come in plan slots, 1, 3 or 10 from Starter to Scale, made in its Studio from a few minutes of training video plus a consent recording, and its pricing page lists avatars from one image or a text prompt on Scale. We charge $199 one time per custom avatar, with the presenter's consent, and it works in the API and in Voice Agents. On latency, Beyond Presence quotes 250 ms from audio input to avatar video, and our avatar starts speaking less than 250 ms after your agent's audio reaches it, measured on our servers: treat the two as level. Beyond Presence stores customer data in Europe, and Enterprise adds zero data retention, region pinning, isolated or on-premise deployments, SLAs and a data processing agreement.

If your own agent drives Beyond Presence through Speech-to-Video (the LiveKit plugin or Pipecat), switching changes only the face. If Managed Agents run the conversation, rebuild that part first, in your own stack or in Voice Agents.

1. List what each Beyond Presence agent or session uses: Speech-to-Video or Managed Agents, custom avatars, camera vision, the knowledge base and meeting joins.
2. If Managed Agents run it, move the prompt, LLM and voice to your own stack (OpenAI Realtime, LiveKit Agents or Pipecat), or rebuild the prompt and tools in Voice Agents if you want it hosted.
3. Create a Talking Avatar Realistic Avatars account (it comes with $1 of free credits), create an API key and pick a ready-made avatar, or order a custom one for $199, one time, with the presenter's consent.
4. Replace the Beyond Presence avatar session in your LiveKit worker or Pipecat pipeline with a Talking Avatar session that takes your agent's audio over WebRTC or in your LiveKit room.
5. Run both on a slice of traffic and compare the cost per conversation in one currency, at your bank's exchange rate, before you cancel the plan.