Compare realtime APIs

Pick up to four APIs from any category and see them side by side.

Unmute
Kyutai
Moshi
Kyutai
PersonaPlex-7B
NVIDIA
CategoryOpen modelsOpen modelsOpen models
StatusGAGAGA
Est. per minuten/an/an/a
How that was worked outNo licence fee; you pay for GPU time. Plus LLM cost if you use a paid LLM endpoint.No licence fee; you pay for GPU time. No licence fee; you pay for GPU time.
Pricing modelfreefreefree
Free tierOpen weightsOpen weightsOpen weights
Connects byWebSocket (OpenAI-Realtime-like backend), Web frontendWebSocket (built-in web server and client)WebSocket (moshi.server web UI)
Audio inBrowser mic via the bundled frontend24 kHz via the Mimi codec (12.5 Hz frames, 80 ms)24 kHz
Audio outStreamed TTS audio24 kHz Mimi-decoded speech24 kHz
LanguagesEnglish and French for STT/TTS (per model names); LLM language depends on your model.English (not stated on the repo page; the released models were trained for English).English.
Latency (vendor claim)Vendor: TTS latency about 750 ms on a single L40S, about 450 ms with services on separate GPUs.Vendor claim: 160 ms theoretical, as low as 200 ms in practice on an L4 GPU.Vendor FullDuplexBench: smooth turn-taking latency 0.170, user-interruption latency 0.240 (units not given on the card).
Key limits
  • No built-in tool calling (wrap your LLM server to add it)
  • Linux x86_64 only (Windows via WSL); no aarch64 or native Mac
  • Small 7B text backbone: limited knowledge and reasoning
  • No built-in tool/function calling
  • Context length not stated on the repo
  • English only
  • No documented tool calling (community forks exist)
High-severity warnings
  • None
  • Not an agent brain
  • None
ComplianceSelf-hosted; your responsibility.Self-hosted; compliance is your responsibility.Self-hosted.
Self-hostableYesYesYes
Last checked2026-10-102026-10-102026-10-10
Key numbers and features
TypeVoice-to-voiceVoice-to-voiceVoice-to-voice
Params B-7.78
VRAM GB162424
CPU okNoNoNo
LicenceCC-BY-4.0CC-BY-4.0Custom
CommercialYesYesYes
Full duplexNoYesYes
StreamingYesYesYes
Latency ms750200-
Languages211
High warnings010