AI project total cost

The calculator vendors don't give you. Token prices are the headline; the bill is the stack. Fill what applies to your project, leave the rest at zero.

Model (for LLM inference)

Prices below update daily. Full table →

LLM input tokens / month (millions)

≈ 15M words of context fed to the model

LLM output tokens / month (millions)

Voice minutes / month

Speech + telephony, $/min combined below

Voice rate $/min (STT+TTS+telephony)

Images / month

Price per image $

Vector database $/month

Fixed infra $/month (hosting, queues, storage)

Evals & observability (% of LLM spend)

Tracing, logging, eval runs — industry default ≈10%

Retries & fallbacks (% of LLM spend)

Failed calls re-run, model fallback on errors

The formula

llm = inTok × P_in + outTok × P_out (P from today's price table) voice = minutes × rate (STT + TTS + telephony) images = count × price_per_image overhead = (evals% + retries%) × (llm + voice + images) total = llm + voice + images + vectorDB + overhead + fixed

Token prices: daily automatic pull from OpenRouter's public API. Voice default bundles hosted STT (~$0.005/min), neural TTS (~$0.015/min) and US inbound telephony (~$0.015/min) — edit to your vendors. Data as of . Prompt-caching and batch discounts deliberately excluded (conservative case).