Amazon Bedrock pricing calculator: the formula, token presets and add-on prices.
To estimate a Bedrock bill, multiply your monthly input and output tokens by your model's per-token rates from the AWS pricing page, then add any add-ons such as Guardrails, knowledge bases or Flows. Below: the formula, ready-made token volumes for a chatbot, a RAG app and a batch job, the add-on prices checked today, and the levers that cut the total.
- Monthly cost for a text model = (input tokens per month ÷ 1,000,000 × input price per 1M) + (output tokens per month ÷ 1,000,000 × output price per 1M). Take the two prices for your model and Region from the AWS Bedrock pricing page. Image, video and speech models bill per image, second or other unit instead.
- Then add what sits around the model. Prices on the pricing page on 10 October 2026: Guardrails content filters $0.15 per 1,000 text units; managed knowledge base storage $5.00 per GB per month and retrieval $1.00 per 1,000 calls; Flows $0.035 per 1,000 node transitions; prompt routing $1 per 1,000 requests. Batch inference was 50% below on-demand for select models.
- The model you pick moves the total most. Batch, prompt caching and routing come next. If a partner builds the workload, the partner can file POC or IW funding when it qualifies and AWS approves.
IThe Bedrock cost formula
Every estimate reduces to one line of arithmetic.
Text models bill per token, separately for input and output. A token is roughly three quarters of an English word, so 1,000 tokens is about 750 words. Input is everything you send: system prompt, history or retrieved passages, and the user's message. Output is what the model writes. Output usually costs several times more per token than input.
Monthly cost = (monthly input tokens ÷ 1M × input price per 1M) + (monthly output tokens ÷ 1M × output price per 1M), summed over every model you call. Get both prices for your model, Region and tier from the Bedrock pricing page. The general AWS Pricing Calculator is useful for the rest of your stack; check whether it covers your specific Bedrock model before you rely on it.
- Monthly input tokens: Input tokens per request times requests per month.
- Monthly output tokens: Typical answer length in tokens times requests per month. Set max_tokens close to it; a high value also eats quota.
- Prices: Per 1M tokens for your model and Region. Some pages quote per 1,000 tokens; multiply by 1,000 to compare.
IIToken presets for three common workloads
Pick the row closest to your app, then multiply by your model's two prices.
The token sizes below are planning assumptions, not measurements: a chatbot with a 1,000-token system prompt plus history, a RAG app that sends about 3,000 tokens of retrieved passages, and a batch job that classifies or summarizes 1,000-token items. Replace them with your own numbers once you have logs.
| Preset | Per request (in / out) | Requests per month | Monthly input | Monthly output |
|---|---|---|---|---|
| Chatbot, small | 1,500 / 300 | 30,000 | 45M | 9M |
| Chatbot, medium | 1,500 / 300 | 300,000 | 450M | 90M |
| Chatbot, large | 1,500 / 300 | 3,000,000 | 4,500M | 900M |
| RAG app, small | 3,500 / 400 | 20,000 | 70M | 8M |
| RAG app, medium | 3,500 / 400 | 200,000 | 700M | 80M |
| Batch job | 1,000 / 150 | 1,000,000 items | 1,000M | 150M |
IIIAdd-on prices to include
These sit around the model and are easy to forget.
| Item | Price | Notes |
|---|---|---|
| Guardrails: content filters (text), denied topics | $0.15 per 1,000 text units | A text unit is up to 1,000 characters |
| Guardrails: sensitive information filters, contextual grounding | $0.10 per 1,000 text units | Regex-based filters and word filters are free |
| Managed knowledge base storage | $5.00 per GB of raw data per month | Managed parsing, embeddings and reranking included |
| Managed knowledge base retrieval | $1.00 per 1,000 standard calls | Agentic retrieval $4.00 per 1,000 calls plus the underlying Retrieve calls |
| Flows | $0.035 per 1,000 node transitions | Plus the model calls and other services the flow uses |
| Intelligent Prompt Routing | $1 per 1,000 requests | Plus the tokens of the model that answers |
| Model evaluation, human review | $0.21 per completed task | Algorithmic scores carry no extra charge |
| Provisioned Throughput | Hourly per model unit | Required for customized models; see the PT page |
IVThe levers that change your bill
In order of size, for most apps.
- Model choice: Prices across the catalog differ by more than an order of magnitude. Use the cheapest model that passes your quality bar, tested with model evaluation.
- Batch inference: For work that does not need an instant answer, the pricing page lists select models at 50% below on-demand.
- Prompt caching: Repeated prompt prefixes bill at a lower cache-read rate; in the pricing page example for Claude 3.5 Sonnet v2, cache reads cost a tenth of the input price. See prompt caching.
- Routing: Send easy prompts to a cheaper model. See Intelligent Prompt Routing.
- Provisioned Throughput: Cheaper than on-demand only at steady, high volume. See Provisioned Throughput.
VEstimate your own bill in six steps
Write each number down. The estimate is only as good as the token counts.
- Step 1: tokens per request: Count input (system prompt, history or passages, user message) and typical output. Use the model's tokenizer or Bedrock's token counting API on real examples.
- Step 2: requests per month: Active users times requests per user per day times 30, or items per batch run times runs per month. Plan for peaks and growth.
- Step 3: prices: Read the input and output prices per 1M tokens for your model and Region from the pricing page.
- Step 4: apply the formula: Monthly input ÷ 1M × input price, plus monthly output ÷ 1M × output price.
- Step 5: apply the levers: Halve batch-eligible work. Reduce the cached share of input to the cache-read rate. Move easy traffic to a cheaper model.
- Step 6: add the add-ons: Guardrails, knowledge base storage and retrieval or your vector store, Flows, routing fees, and S3 and CloudWatch.
VIWho can file AWS funding for a Bedrock build
There are two kinds of AWS money for a Bedrock build. A partner files one. You apply for the other yourself.
Partner-filed funding. An AWS consulting partner can file POC (proof of concept) funding to test a Bedrock workload, whether or not you run on AWS today. If you already run on AWS and are adding new AI work, the partner can file incremental workload (IW) funding instead. For startups this is often $50K to $150K. Depends on the workload; AWS approves. The work has to be real, and the partner has to be the one doing it.
AWS Activate, which you apply for yourself. Activate Founders gives $1,000 to start, up to $5,000, with no investor needed. Activate Portfolio gives up to $200,000 if your VC, accelerator or angel group is an Activate Provider and gives you its Org ID. A consulting partner cannot apply for Activate on your behalf. AWS made Activate credits usable for third-party models on Bedrock in April 2024.
For most startups without a VC, the partner route is the larger of the two. Your estimate from this page is a good input for that conversation. CloudRoute matches you with one vetted AWS partner for free. Partners pay CloudRoute a monthly membership plus 10% of AWS-approved funding.
| Funding | Who files it | Amount | When it fits |
|---|---|---|---|
| Proof of concept (POC) funding | Your matched AWS partner | $50K to $150K for startups. Depends on the workload; AWS approves. | You want to test something, whether or not you are on AWS today. |
| Incremental workload (IW) funding | Your matched AWS partner | $50K to $150K for startups. Depends on the workload; AWS approves. | You are already on AWS and are launching new work: a new AI feature, a new customer or a new contract. |
| AWS Activate Founders | Founder applies | $1,000 to start, up to $5,000 | Self-funded startups with no investor or accelerator behind them. |
| AWS Activate Portfolio | Founder applies | up to $200,000 | Startups backed by a VC firm, accelerator or angel investor that is an AWS Activate Provider. |
| AWS Credits for AI Startups | Invite only (AWS account manager) | $200,000 or more, by invitation | AI startups ready to scale after Activate Portfolio, invited by AWS. |
How a match works (example)
- You tell us about the work
You fill in a short form: what you run today, what you want to move, test or build, your stage and your region.
- We pick one partner
We read your answers and pick one vetted AWS consulting partner that fits the work, the stack and the region. We reply within 24 hours; your matched partner reaches out within 5 business days.
- The partner checks POC funding
Your matched partner scopes the test with you. Then it checks whether the test qualifies for POC (proof of concept) funding. A POC can fit whether or not you run on AWS today.
- The partner files, AWS decides
If the work qualifies, the partner files the funding request with AWS. AWS reviews it and decides. The amount depends on the workload.
- You pay nothing
CloudRoute is free for startups. Partners pay CloudRoute a monthly membership plus 10% of AWS-approved funding.
AWS Activate credits are separate, and you apply for them yourself. Use Activate Founders if no investor backs you. Use Activate Portfolio with the Org ID or invite from your VC, accelerator, incubator or angel group if it is an Activate Provider. A consulting partner cannot apply for Activate on your behalf.
Common questions
Is there an official Amazon Bedrock pricing calculator?
What is the formula for Bedrock cost?
How much does a Bedrock chatbot cost per month?
How do I estimate a RAG app on Bedrock?
How much can batch inference and prompt caching save?
Why does this page not list model prices?
Can AWS funding pay for a Bedrock build?
Get a partner who can build it and file the funding
Tell us what you are building on Bedrock. We match you with one vetted AWS partner. If the work qualifies, the partner files POC funding for a test, or incremental workload (IW) funding for new work on AWS: often $50K to $150K for startups. Depends on the workload; AWS approves. CloudRoute is free for startups.