Eleven v4 Turbo API Pricing: $0.011 per 1K Characters

Eleven v4 Turbo is the low-latency variant of Eleven v4, both launched 28 September 2026. The model id is eleven_v4_turbo. Eleven v4 Turbo API pricing is $0.011 per 1,000 characters, struck through against a list price of $0.04 per 1,000 characters.

Eleven v4 Turbo is not on RouterPlex yet. Register to use the current 60+ models on one prepaid key. The new speech model is available here once the marketplace lists it. Today's chat routes do not return audio.

Eleven v4 Turbo API pricing #

The Turbo card uses the same 72% off until Oct 12 label as v4. The page does not print the year. $0.011 / $0.04 matches that discount. Per 1M characters the promo is $11 and the list is $40. One thousand characters is $0.011 on promo and $0.04 at list. The page also prints about $0.01 per minute at the promo rate.

That promo is ElevenLabs' price. It is not a RouterPlex credit and it is not a rate this catalog bills.

Latency: 100 ms and 150 ms are different measurements #

The docs describe eleven_v4_turbo as median inference latency of about 100 ms. The dagger says that figure excludes application and network latency.

The launch chart is a different clock. It labels median time to first speech:

Measured September 2026, identical scripts, default settings, Turbo over WebSocket. The chart's bars have two segments, audio arrives and silence before speech starts. The totals above are the labeled values; ElevenLabs does not print the split between the two segments.

The blog footnote also says network latency was measured and removed. Quote the chart total when you compare vendors. Quote the docs' 100 ms figure when you mean inference time with application and network excluded.

How you call it #

v4 itself is the Text to Dialogue HTTP API, with a printed cap of 10,000 characters. Turbo is the websocket path for agents that need the first audio packet quickly. ElevenLabs says both are in ElevenAgents, ElevenCreative, and ElevenAPI.

A live RouterPlex call is still text #

gemini-3.8-flash is live here at $0.75 input and $3.75 output per 1M tokens. It is the chat model, not Gemini 3.8 Flash TTS.

curl https://api.routerplex.com/v1/chat/completions \
  -H "Authorization: Bearer $ROUTERPLEX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-3.8-flash",
    "messages": [
      {"role": "user", "content": "Write a one-sentence hold message for a voice agent."}
    ]
  }'

Balance never goes negative. The hard spend limit is on the key. See the catalog and how the OpenAI-compatible API works.

On RouterPlex, once Turbo is listed #

The route name lands with the catalog row. Price will follow the vendor list price at 0% markup, prepaid, with the same hard per-key budget. Create an account for the 60+ models that are live now, and for eleven_v4_turbo when this marketplace carries it. Until that row exists, the websocket host is ElevenLabs, not api.routerplex.com.