Tasks and inputs
Eleven Realtime v1 Max is listed for Audio, Audio-to-audio. It accepts audio, text and produces audio, text. Use the publisher’s input format and preprocessing for your chosen deployment.
SyncDev model directory · Profile updated Nov 17, 2026
Eleven Realtime v1 Max is ElevenLabs' reasoning speech-to-speech agent for support and multi-step tool workflows, OpenAI Realtime-compatible.
The specification score summarizes recorded features and limits. It is not a benchmark result.
Pricing conditions: About $0.08 per minute plus LLM usage. Indicative; confirm on your ElevenLabs plan.
Recorded modality support shows what this model can accept and produce. An undocumented modality is shown as unknown rather than assumed to be unsupported.
Accepts
Produces
Customer support calls are rarely one question long. Look up the order, check the refund policy, confirm the address, book the courier. Eleven Realtime v1 Max is ElevenLabs' voice-agent configuration for that kind of call. It reasons before it answers, it handles complex instructions, and it is built for multi-step tool calling in the middle of a live conversation.
Like its sibling, it is a tuned bundle rather than a single network: speech recognition, turn-taking, an undisclosed LLM and speech output, delivered as one speech-to-speech system.
Point three is easy to miss. Teams who toggled Realtime mode expecting the lighter configuration may be on Max, with different latency and cost.
eleven_realtime_v1_maxElevenLabs quotes 99% on a speech-reasoning measure and 570 ms time to first audio for Max. Interestingly, that latency figure is lower than the one quoted for the lighter Mini configuration. These are vendor measurements under unstated conditions; do not assume Max is faster in your deployment without testing. The LLM token cost on top of the per-minute rate is likely to be the larger variable, because reasoning consumes tokens.
Use Max where a wrong answer costs money: support, lead qualification, booking and other flows that touch your systems. Use Mini where tone and fluidity matter more than accuracy, such as companions and tutors. Switching between them is a model selection, not an integration change, so an A/B test is cheap.
Take a dozen real support tickets that needed two or more system lookups to resolve. Turn them into call scripts, connect Max to your actual tools in a staging environment and run each script five times. Score task completion, wrong tool arguments and how often a human would have had to step in. Record end-of-speech to first-audio latency on every turn and the total LLM spend per call.
For comparison, run the same scripts on Gemini 3.8 Live or GPT-Realtime-2.1. Other options are listed under audio-to-audio models.
Compare the model on the work you need to complete.
Eleven Realtime v1 Max is listed for Audio, Audio-to-audio. It accepts audio, text and produces audio, text. Use the publisher’s input format and preprocessing for your chosen deployment.
Compare pronunciation, voice consistency and time to the first playable audio. Check the service’s billing unit for the actual script.
Use the model identifier and account requirements for the selected API route. Confirm its request limits and output format with one complete task before connecting a larger workflow.
These records share a directory category. Check the source card, output type and deployment requirements before treating a model as a replacement.
rime
Rime Prism is a speech-to-speech model for enterprise voice agents that settles each reply as inspectable text first, with parallel tool calls.
Liquid AI
LFM3-Audio-2B is Liquid AI's open 2B speech model: ASR, TTS, voice chat and interleaved text-and-speech output from one local checkpoint.
Alibaba
Qwen-Audio-3.1-Realtime Flash is Alibaba's lighter full-duplex speech model with audio and text in and out, function calling, web search and a 262K context.
ElevenLabs
Eleven Realtime v1 Mini is ElevenLabs' fast, low-cost speech-to-speech voice agent stack, with an OpenAI Realtime-compatible API and 90+ languages.
xAI
Grok Voice 3.0 is xAI's speech-to-speech model for real-time voice agents, replacing Grok Voice 2.0 at the same $0.08-per-minute price.
Amazon
Nova 2.5 Sonic is Amazon's speech-to-speech model on Bedrock for real-time voice agents, with async tool calling, polyglot voices and a 256K context.