Text-to-Speech Cost Calculator

Compare ElevenLabs, OpenAI, Google Cloud, Amazon Polly and Azure text-to-speech costs for your script or monthly character volume.

See the cheapest voice tier, cost per audio hour and which ElevenLabs plan covers your usage.

Last updatedHow we build & check our tools
Results will appear here once you enter your values.

How text-to-speech is billed

Almost every provider bills per character. Google Cloud, Amazon Polly, Azure AI Speech and OpenAI's tts-1 models charge a rate per 1 million characters, and the count includes spaces, punctuation and newlines. Google also counts most SSML tags. The monthly cost is simply characters times the rate: 180,000 characters on OpenAI tts-1 at $15 per 1M is $2.70.

ElevenLabs sells plans and pay-as-you-go. As of September 2026 its API page lists $0.10 per 1,000 characters for Multilingual v2 and v3 voices (that is $100 per 1M) and 0.05 per 1,000 for Flash and Turbo. Plans bundle characters: Starter is6 for 60,000, Creator $22 for 220,000, Pro $99 for 990,000, Scale $299 for 2.99 million and Business $990 for 9.9 million multilingual characters, with usage above the allowance billed at the same per-character rate.

Converting words and minutes. English averages roughly 6 characters per word once you count the trailing space, and a relaxed narration pace is about 150 words per minute. One audio minute is therefore about 900 characters and one hour about 54,000. Both numbers are editable because fast explainer voices can reach 180 words per minute and technical text runs longer per word.

Where the savings are

Standard voices are cheap enough to ignore. Google Standard and WaveNet voices and Amazon Polly Standard all cost $4 per 1M characters, so a full hour of narration costs about $0.22. Google gives 4 million free characters a month across Standard and WaveNet, and Polly gives 5 million Standard characters a month, which covers roughly 74 to 92 hours of audio.

Mid-tier neural voices sit between $15 and $30. Azure Neural is $15, Google Neural2 and Polly Neural are $16, and Google Chirp 3 HD, Polly Generative and OpenAI tts-1-hd are $30 per 1M. At 54,000 characters per hour that is $0.81 to $1.62 per finished hour.

Premium voices cost several times more. ElevenLabs multilingual works out to about 5.40 per hour at list overage rates, Polly Long-Form is100 per 1M and Google Studio is $160 per 1M, or $8.64 per hour.

Caveats before you commit

Prices change often. Every rate in this calculator was read from the provider's official pricing page in September 2026 and each table row links to its source. Check the provider before buying, and override any rate in the table if you have a negotiated or regional price.

Free tiers are not equal. Polly's Neural, Long-Form and Generative allowances only apply in your first 12 months, Azure's 0.5 million free characters belong to the F0 tier, and the ElevenLabs free plan includes only 10,000 multilingual characters. Leave the free tier box unchecked for a steady-state budget.

Retakes cost money too. Regenerating a paragraph to fix a mispronunciation bills the characters again, so add 10 to 20 percent for revisions on long-form work. OpenAI gpt-4o-mini-tts is billed per text and audio token and is not included here.

Worked example: a weekly podcast

Inputs. Four 25-minute episodes a month at 150 words per minute is 15,000 words, or about 90,000 characters at 6 characters per word.

Results. Google Standard or Polly Standard costs 0.36 before the free tier and nothing after it. OpenAI tts-1 costs1.35, Google Chirp 3 HD $2.70 before its 1 million free characters, and ElevenLabs pay-as-you-go multilingual $8.00 after the 10,000 free characters. The ElevenLabs plan that covers 90,000 characters is Creator at 22, but pay-as-you-go at8.00 is cheaper than Creator, and Starter plus overage is $9.00, so the calculator shows both the covering plan and the cheapest total.

Frequently Asked Questions

Common questions about the Text-to-Speech Cost Calculator

As of September 2026 the ElevenLabs API pricing page lists $0.10 per 1,000 characters for Multilingual v2 and v3 voices and $0.05 per 1,000 for Flash and Turbo. Plans include characters: Starter $6 for 60,000, Creator $22 for 220,000, Pro $99 for 990,000. Check the ElevenLabs site before buying because plans change.
From the same team

Stop paying per token — route AI requests to your own GPU

Wide Area AI is a local-first AI gateway: repeated requests hit an edge cache, the rest run free on your own hardware, and the cloud is only a failover. OpenAI-compatible endpoint, free tier.

Start routing — free
VRAM CalculatorCan I Run This AI?AI API Cost CalculatorGPU for Model

Sources & References

ElevenLabs API Pricing

Per-1K character TTS rates and characters included in each plan, September 2026.
→Source: ElevenLabs

API Pricing

tts-1 at $15 and tts-1-hd at $30 per 1M characters.
→Source: OpenAI

Text-to-Speech pricing

Standard, WaveNet, Neural2, Chirp 3 HD and Studio rates, free tiers and character counting rules.
→Source: Google Cloud

Amazon Polly Pricing

Standard, Neural, Long-Form and Generative per-1M rates and free tier allowances.
→Source: Amazon Web Services

Azure Speech pricing

Neural ($15) and Neural HD ($22) per 1M characters via the Azure Retail Prices API, and the 0.5M free F0 allowance.
→Source: Microsoft Azure

Explore More Tools

Explore these related tools for your next step.

ai

LLM Token Usage Calculator

Calculate token usage and associated costs for LLM prompts.

Try it now
ai

AI Embedding Cost Calculator

Estimate the total token count and API cost to embed a document corpus using OpenAI, Cohere, or Voyage embedding models.

Try it now
ai

Neural Network Playground

Train a real neural network in your browser — no code, no installs. Pick a sample dataset (Titanic survival, vehicle classifier, handwritten digits) or upload your own CSV, watch the network learn with live visuals, then ask it questions and export the model as Python/Keras code.

Try it now
ai

LLM Token Counter

Count tokens in text for GPT, Claude, Gemini, and open models like Llama, Gemma, and Qwen — or search any model on Hugging Face. Estimate API costs, check context window fit, and optimize prompts

Try it now
ai

LLM VRAM Calculator

Calculate how much VRAM any LLM needs to run locally — pick a model (Llama, Gemma, Qwen, DeepSeek, or search Hugging Face), choose a quantization and context size, and see which GPUs it fits on, including multi-GPU setups.

Try it now
ai

LLM Inference Speed Calculator

Estimate LLM tokens per second from memory bandwidth, model size, quantization, and context window. Compare generation speed across GPUs (RTX 4090, 5090, A100, H100, Apple Silicon) and understand the memory-bandwidth bottleneck.

Try it now