Get agent estimates

Returns the estimate the agent builder shows in its header: cost per minute by product, an end-to-end latency range and the prompt size in tokens. Values come from the pricing and latency tables in packages/shared. Beyond Retell

Request#

Path parameters#

Query parameters#

Example request#

curl "https://api.tried.com.au/agent-estimates/agent_3f9a1c0b2d4e5f6a7b8c9d0e" \
  -H "Authorization: Bearer $TRIED_API_KEY"

const res = await fetch("https://api.tried.com.au/agent-estimates/agent_3f9a1c0b2d4e5f6a7b8c9d0e", {
  headers: { Authorization: `Bearer ${process.env.TRIED_API_KEY}` },
});
const data = await res.json();

import os, requests

res = requests.get(
    "https://api.tried.com.au/agent-estimates/agent_3f9a1c0b2d4e5f6a7b8c9d0e",
    headers={"Authorization": f"Bearer {os.environ['TRIED_API_KEY']}"},
)
data = res.json()

Response#

200 OK

{
  "cost_per_min": 0.3,
  "cost_breakdown": [
    { "product": "Voice infra", "per_min": 0.08 },
    { "product": "LLM (gpt-5.6-terra)", "per_min": 0.1 },
    { "product": "Voice (Premium)", "per_min": 0.12 }
  ],
  "latency_ms": { "min": 820, "max": 1290, "breakdown": { "tts": 190, "llm": [500, 800], "stt": [130, 300] } },
  "prompt_tokens": { "min": 420, "max": 2520 }
}

latency_ms is the sum of the TTS, LLM and STT figures for the agent's voice_model, model and stt_mode. prompt_tokens.min is the prompt characters divided by 4, rounded up, and max adds 2100. cost_breakdown adds Knowledge base when the engine has knowledge bases, and an STT line for stt_mode: "custom" with a provider other than Deepgram. For a speech-to-speech LLM (s2s_model), the LLM and TTS lines are replaced by one speech-to-speech line.

Errors#

See Errors for the body format.