BroadRouter

Amazon Bedrock API pricing

Updated

Amazon Bedrock lists standard per-token prices for 108 models. The cheapest is Voxtral Mini 1.0 at USD 0.04 input and USD 0.04 output per 1M tokens; the dearest is Claude Opus 4.1 at USD 15 / USD 75. It publishes cached-input prices for 34 of them, plus batch, flex / spot, priority and fast tiers. Checked 2026-11-15.

Amazon Bedrock hosts models it does not train. By maker: Anthropic 26, Meta 14, OpenAI 14, Mistral AI 13, Google 6, AI21 Labs 5, and 9 more. 146 prices, each read from Amazon Bedrock's own pricing page or one of 13 linked model pages, checked 2026-11-15. BroadRouter is BroadNet's own site and does not sell Amazon Bedrock's services. Automated re-check 2026-11-16: 146 of 146 prices still matched the provider's page (4 of them on pages that load prices with JavaScript, re-read with a page renderer within the last 7 days).

Models with a standard price
108
Median input / output
USD 0.72 / USD 2.30
Cached-input prices
34 of 108
Cheapest host for
7 of 68 shared models

Which models is Amazon Bedrock the cheapest provider for?

Of the 68 models that Amazon Bedrock and at least one other provider both price at the standard tier, Amazon Bedrock has the lowest price for 7, ranked on a blend of 3 input tokens to 1 output token (1 of them tied with another provider at the same price). It also ties the lowest price for 4 more, where another provider is listed first. It is the dearest for 35. Another 40 models have a standard price from Amazon Bedrock only.

Models where Amazon Bedrock is the cheapest provider
ModelAmazon Bedrock input / outputNext cheapestSaving on 3:1 blendProviders
GPT OSS Safeguard 20BUSD 0.07 / USD 0.20Featherless USD 0.075 / USD 0.3022%3
Llama 3.2 11B Vision InstructUSD 0.16 / USD 0.16Cloudflare Workers AI USD 0.049 / USD 0.67622%3
Mistral Large 3USD 0.50 / USD 1.50IBM watsonx.ai USD 0.636 / USD 1.90821%3
Gemma 3 4BUSD 0.04 / USD 0.08DeepInfra USD 0.05 / USD 0.1020%3
Llama 3.1 Instruct (70B)USD 0.72 / USD 0.72Cloudflare Workers AI USD 0.293 / USD 2.2538%3
Minimax M2.1USD 0.30 / USD 1.20Featherless USD 0.30 / USD 1.20tie3
Llama 3 Instruct (70B)USD 2.65 / USD 3.50Featherless USD 3.50 / USD 8.3039%2

The 6 models Amazon Bedrock sells that the most other providers also sell, with Amazon Bedrock's rank among them on the 3:1 blend.

Amazon Bedrock on the most widely hosted models
ModelAmazon BedrockRankCheapest
GLM 5.3USD 1.848 / USD 5.80823 of 24DeepInfra USD 0.5625 / USD 2.50
OpenAI GPT OSS 120BUSD 0.15 / USD 0.6011 of 21CoreWeave Forge USD 0.03 / USD 0.17
Kimi K3USD 3.30 / USD 16.5017 of 18Chutes USD 1.85 / USD 9.50
Llama 3.3 70BUSD 0.72 / USD 0.729 of 14DeepInfra USD 0.10 / USD 0.32
Gemma 4 31BUSD 0.14 / USD 0.404 of 12CoreWeave Forge USD 0.10 / USD 0.34
Gemma 4 26B A4BUSD 0.13 / USD 0.405 of 11DeepInfra USD 0.07 / USD 0.34

Which model makers does Amazon Bedrock sell?

Amazon Bedrock prices models from 15 makers. Anthropic leads with 26 models. Standard prices, USD per 1M input / output tokens.

Amazon Bedrock models by maker
MakerModelsCheapest modelMedian input / output
Anthropic26Claude 3 Haiku USD 0.25 / USD 1.25USD 3.85 / USD 19.25
Meta14Llama 3.2 1B Instruct USD 0.10 / USD 0.10USD 0.30 / USD 0.72
OpenAI14GPT OSS Safeguard 20B USD 0.07 / USD 0.20USD 1.65 / USD 8.25
Mistral AI13Voxtral Mini 1.0 USD 0.04 / USD 0.04USD 0.40 / USD 0.70
Google6Gemma 3 4B USD 0.04 / USD 0.08USD 0.11 / USD 0.335
AI21 Labs5Jamba 1.5 Mini USD 0.20 / USD 0.40USD 2 / USD 8
Qwen5Qwen3 32B FP8 USD 0.15 / USD 0.60USD 0.15 / USD 1.20
Cohere4Command-Light USD 0.30 / USD 0.60USD 0.75 / USD 1.75
Moonshot AI4Kimi K2 Thinking USD 0.60 / USD 2.50USD 0.60 / USD 3
NVIDIA4NVIDIA Nemotron Nano 2 USD 0.06 / USD 0.23USD 0.105 / USD 0.42
Z.ai4GLM 4.7 Flash USD 0.07 / USD 0.40USD 0.80 / USD 2.70
MiniMax3MiniMax M2 USD 0.30 / USD 1.20USD 0.30 / USD 1.20
Writer3Palmyra Vision 7B USD 0.15 / USD 0.60USD 0.60 / USD 6
xAI3Grok 4.3 USD 1.25 / USD 2.50USD 2.20 / USD 6.60
DeepSeek2DeepSeek V3.2 USD 0.62 / USD 1.85USD 0.985 / USD 3.625

Amazon Bedrock price list

USD per 1M tokens: 146 prices for 110 models, one row per model with any other service tiers stacked under the standard price. Sources checked 2026-11-15 unless a row says otherwise. "vs maker" compares a standard price with the model maker's own API on the 3:1 blend; a minus means Amazon Bedrock is cheaper. Of 34 hosted model prices with a maker price to compare, 0 are below the maker's and 28 above.

Amazon Bedrock API prices by model
ModelTierInputCachedOutputContextvs makerSource
Claude 2.0standardUSD 8—USD 24not stated—source 2026-11-14
Claude 2.1standardUSD 8—USD 24not stated—source 2026-11-14
Claude 3 Haikustandard
batch
USD 0.25
USD 0.125
—
—
USD 1.25
USD 0.625
not stated—source 2026-11-15, 2026-11-14
Claude 3 Opusstandard
batch
USD 15
USD 7.50
—
—
USD 75
USD 37.50
not stated—source 2026-11-14, 2026-11-15
Claude 3 Sonnetstandard
batch
USD 3
USD 1.50
—
—
USD 15
USD 7.50
not stated—source 2026-11-14
Claude 3.5 Haikustandard
batch
USD 0.80
USD 0.40
USD 0.08
—
USD 4
USD 2
not stated—source 2026-11-14
Claude 3.5 Sonnetstandard
batch
USD 3
USD 1.50
—
—
USD 15
USD 7.50
not stated—source 2026-11-14
Claude 3.5 Sonnet v2standard
batch
USD 3
USD 1.50
USD 0.30
—
USD 15
USD 7.50
not stated—source 2026-11-14
Claude 3.7 Sonnetstandard
batch
USD 3
USD 1.50
USD 0.30
—
USD 15
USD 7.50
not stated—source 2026-11-14
Claude Fable 5standardUSD 11USD 1.10USD 55not stated+10%source 2026-11-14
Claude Fable 5.1standardUSD 11USD 0.275USD 55not stated+10%source 2026-11-14
Claude Haiku 4.5standard
batch
USD 1.10
USD 0.55
USD 0.11
—
USD 5.50
USD 2.75
not stated+10%source 2026-11-14
Claude InstantstandardUSD 0.80—USD 2.40not stated—source 2026-11-14
Claude Opus 4standard
batch
USD 15
USD 7.50
USD 1.50
—
USD 75
USD 37.50
not stated—source 2026-11-14, 2026-11-15
Claude Opus 4.1standardUSD 15USD 1.50USD 75not stated—source 2026-11-14
Claude Opus 4.5standard
batch
USD 5.50
USD 2.75
USD 0.55
—
USD 27.50
USD 13.75
not stated+10%source 2026-11-14
Claude Opus 4.6standard
batch
USD 5.50
USD 2.75
USD 0.55
—
USD 27.50
USD 13.75
not stated+10%source 2026-11-14
Claude Opus 4.7standardUSD 5.50USD 0.55USD 27.50not stated+10%source 2026-11-14
Claude Opus 4.8standardUSD 5.50USD 0.55USD 27.50not stated+10%source 2026-11-14
Claude Opus 5standard
batch
USD 5.50
USD 2.75
USD 0.55
—
USD 27.50
USD 13.75
not stated+10%source 2026-11-14
Claude Opus 5.5standardUSD 4.40USD 0.22USD 22not stated+10%source 2026-11-14
Claude Sonnet 4standard
batch
USD 3
USD 1.50
USD 0.30
—
USD 15
USD 7.50
not stated—source 2026-11-14
Claude Sonnet 4.5standard
batch
USD 3.30
USD 1.65
USD 0.33
—
USD 16.50
USD 8.25
not stated+10%source 2026-11-14
Claude Sonnet 4.6standard
batch
USD 3.30
USD 1.65
USD 0.33
—
USD 16.50
USD 8.25
not stated+10%source 2026-11-14
Claude Sonnet 5standardUSD 2.20USD 0.22USD 11not stated+10%source 2026-11-14
Claude Sonnet 5.5standardUSD 2.20USD 0.11USD 11not stated+10%source 2026-11-14
CommandstandardUSD 1—USD 2not stated—source 2026-11-14
Command RstandardUSD 0.50—USD 1.50not stated+186%source 2026-11-14
Command R+standardUSD 3—USD 15not stated—source 2026-11-14
Command-LightstandardUSD 0.30—USD 0.60not stated—source 2026-11-14
DeepSeek R1standardUSD 1.35—USD 5.40not stated—source 2026-11-14
DeepSeek V3.2standardUSD 0.62—USD 1.85not stated—source 2026-11-14
Devstral 2 123BstandardUSD 0.40—USD 2not stated—source 2026-11-14
Gemma 3 12BstandardUSD 0.09—USD 0.29not stated—source 2026-11-14
Gemma 3 27BstandardUSD 0.23—USD 0.38not stated—source 2026-11-14
Gemma 3 4BstandardUSD 0.04—USD 0.08not stated—source 2026-11-14
Gemma 4 26B A4BstandardUSD 0.13—USD 0.40not stated—source 2026-11-14
Gemma 4 31BstandardUSD 0.14—USD 0.40not stated—source 2026-11-14
Gemma 4 E2BstandardUSD 0.04—USD 0.08not stated—source 2026-11-14
GLM 4.7standardUSD 0.60—USD 2.20not statedsamesource 2026-11-14
GLM 4.7 FlashstandardUSD 0.07—USD 0.40not stated—source 2026-11-14
GLM 5standardUSD 1—USD 3.20not statedsamesource 2026-11-14
GLM 5.3standardUSD 1.848USD 0.3432USD 5.808not stated+32%source 2026-11-14
GPT OSS 20Bstandard
batch
flex / spot
priority
USD 0.07
USD 0.035
USD 0.035
USD 0.1225
—
—
—
—
USD 0.30
USD 0.15
USD 0.15
USD 0.525
not stated—source 2026-11-14
GPT OSS Safeguard 120BstandardUSD 0.15—USD 0.60not stated—source 2026-11-14
GPT OSS Safeguard 20BstandardUSD 0.07—USD 0.20not stated—source 2026-11-14
GPT-5.4standardUSD 2.75USD 0.275USD 16.501,050,000+10%source 2026-11-14
GPT-5.5standardUSD 5.50USD 0.55USD 331,050,000+10%source 2026-11-14
GPT-5.6 LunastandardUSD 0.22USD 0.022USD 1.321,050,000+10%source 2026-11-14
GPT-5.6 SolstandardUSD 4.40USD 0.44USD 221,050,000+10%source 2026-11-14
GPT-5.6 TerrastandardUSD 2.20USD 0.22USD 13.201,050,000+10%source 2026-11-14
GPT-6 Astrastandard
fast
USD 11
USD 66
USD 1.10
USD 6.60
USD 55
USD 330
1,050,000+10%source 2026-11-14
GPT-6 LunastandardUSD 0.11USD 0.011USD 0.551,050,000+10%source 2026-11-14
GPT-6 SolstandardUSD 2.20USD 0.22USD 111,050,000+10%source 2026-11-14
GPT-6.1 Solstandard
fast
USD 2.20
USD 13.20
USD 0.11
USD 0.66
USD 11
USD 66
1,050,000+10%source 2026-11-14
GPT-6.1 TerrastandardUSD 1.10USD 0.055USD 5.501,050,000+10%source 2026-11-14
Grok 4.3standardUSD 1.25USD 0.20USD 2.501,000,000samesource 2026-11-14
Grok 4.6standardUSD 2.20USD 0.55USD 6.60500,000+10%source 2026-11-14
Grok 4.7standardUSD 2.20USD 0.55USD 6.60524,288+10%source 2026-11-14
Jamba 1.5 LargestandardUSD 2—USD 8not stated—source 2026-11-14
Jamba 1.5 MinistandardUSD 0.20—USD 0.40not stated—source 2026-11-14
Jamba-InstructstandardUSD 0.50—USD 0.70not stated—source 2026-11-14
Jurassic-2 MidstandardUSD 12.50—USD 12.50not stated—source 2026-11-14
Jurassic-2 UltrastandardUSD 18.80—USD 18.80not stated—source
Kimi K2 ThinkingstandardUSD 0.60—USD 2.50not stated—source 2026-11-14
Kimi K2.5standardUSD 0.60—USD 3not stated—source 2026-11-14
Kimi K3standardUSD 3.30USD 0.33USD 16.50not stated+10%source 2026-11-14
Kimi K3.1 ProfastUSD 4.40USD 0.22USD 22not stated—source 2026-11-14
Llama 2 Chat (13B)standardUSD 0.75—USD 1not stated—source 2026-11-14
Llama 2 Chat (70B)standardUSD 1.95—USD 2.56not stated—source 2026-11-14
Llama 3 8B InstructstandardUSD 0.30—USD 0.60not stated—source 2026-11-14
Llama 3 Instruct (70B)standardUSD 2.65—USD 3.50not stated—source 2026-11-14
Llama 3.1 8B Instructstandard
batch
USD 0.22
USD 0.11
—
—
USD 0.22
USD 0.11
not stated—source 2026-11-14
Llama 3.1 Instruct (405B)batchUSD 1.20—USD 1.20not stated—source 2026-11-14
Llama 3.1 Instruct (70B)standard
batch
fast
USD 0.72
USD 0.36
USD 0.90
—
—
—
USD 0.72
USD 0.36
USD 0.90
not stated—source 2026-11-14
Llama 3.2 11B Vision InstructstandardUSD 0.16—USD 0.16not stated—source 2026-11-14
Llama 3.2 1B InstructstandardUSD 0.10—USD 0.10not stated—source 2026-11-14
Llama 3.2 3B InstructstandardUSD 0.15—USD 0.15not stated—source 2026-11-14
Llama 3.2 Instruct (90B)standardUSD 0.72—USD 0.72not stated—source 2026-11-14
Llama 3.3 70Bstandard
batch
USD 0.72
USD 0.36
—
—
USD 0.72
USD 0.36
not stated—source 2026-11-14
Llama 4 Maverick 17B 128E Instruct FP8standard
batch
USD 0.24
USD 0.12
—
—
USD 0.97
USD 0.485
not stated—source 2026-11-14
Llama 4 Scout 17B 16E Instructstandard
batch
USD 0.17
USD 0.085
—
—
USD 0.66
USD 0.33
not stated—source 2026-11-14
Magistral Small 1.2standardUSD 0.50—USD 1.50not stated—source 2026-11-14
MiniMax M2standardUSD 0.30—USD 1.20not stated—source 2026-11-14
Minimax M2.1standardUSD 0.30—USD 1.20not stated—source 2026-11-14
MiniMax M2.5standardUSD 0.30—USD 1.20not stated—source 2026-11-14
Ministral 3 14BstandardUSD 0.20—USD 0.20not statedsamesource 2026-11-14
Ministral 3 3BstandardUSD 0.10—USD 0.10not statedsamesource 2026-11-14
Ministral 3 8BstandardUSD 0.15—USD 0.15not statedsamesource 2026-11-14
Mistral 7BstandardUSD 0.15—USD 0.20not stated—source 2026-11-14
Mistral Large (24.02)standardUSD 4—USD 12not stated—source 2026-11-14
Mistral Large 3standardUSD 0.50—USD 1.50not stated—source 2026-11-14
Mistral Small (24.02)standard
batch
USD 1
USD 0.50
—
—
USD 3
USD 1.50
not stated—source 2026-11-14
Mixtral 8*7BstandardUSD 0.45—USD 0.70not stated—source 2026-11-14
NVIDIA Nemotron 3 Nano 30B A3BstandardUSD 0.06—USD 0.24not stated—source 2026-11-14
NVIDIA Nemotron 3 Super 120B A12BstandardUSD 0.15—USD 0.65not stated—source 2026-11-14
NVIDIA Nemotron Nano 2standardUSD 0.06—USD 0.23not stated—source 2026-11-14
NVIDIA Nemotron Nano 2 VLstandardUSD 0.20—USD 0.60not stated—source 2026-11-14
OpenAI GPT OSS 120Bstandard
batch
flex / spot
priority
USD 0.15
USD 0.075
USD 0.075
USD 0.2625
—
—
—
—
USD 0.60
USD 0.30
USD 0.30
USD 1.05
not stated—source 2026-11-14
Palmyra Vision 7BstandardUSD 0.15—USD 0.60not stated—source 2026-11-14
Palmyra X4standardUSD 2.50—USD 10not stated—source 2026-11-14
Palmyra X5standardUSD 0.60—USD 6not stated—source 2026-11-14
Pixtral Large (25.02)standardUSD 2—USD 6not stated—source 2026-11-14
Qwen3 32B FP8standard
batch
flex / spot
priority
USD 0.15
USD 0.075
USD 0.075
USD 0.2625
—
—
—
—
USD 0.60
USD 0.30
USD 0.30
USD 1.05
not stated—source 2026-11-14
Qwen3 Coder 30B A3Bstandard
batch
flex / spot
priority
USD 0.15
USD 0.075
USD 0.075
USD 0.2625
—
—
—
—
USD 0.60
USD 0.30
USD 0.30
USD 1.05
not stated—source 2026-11-14
Qwen3 Coder NextstandardUSD 0.50—USD 1.20not stated—source 2026-11-14
Qwen3 Next 80B A3B InstructstandardUSD 0.15—USD 1.20not stated—source 2026-11-14
Qwen3 VL 235B A22BstandardUSD 0.53—USD 2.66not stated—source 2026-11-14
Voxtral Mini 1.0standardUSD 0.04—USD 0.04not stated—source 2026-11-14
Voxtral Small 1.0standardUSD 0.10—USD 0.30not stated—source 2026-11-14
  • Cached "—" means Amazon Bedrock's page shows no cache price for that row; budget cached tokens at the input price.
  • Context "not stated": the source does not give it (15 of 146 rows have one).

Report a wrong price or number (email to info@broadnet.ai, with this page's link filled in).

What service tiers does Amazon Bedrock publish?

  • batch (26 rows). Across 25 models with both prices, the median batch price is 50% below standard. Amazon Bedrock row note: “Batch columns of the Geo/in-region table. Shown for region US East (N. Virginia) on the pricing page (values filled from AWS's published price list for us-east-1).”
  • flex / spot (4 rows). Across 4 models with both prices, the median flex / spot price is 50% below standard. Amazon Bedrock row note: “Flex tier pricing table. Shown for region US East (N. Virginia) on the pricing page (values filled from AWS's published price list for us-east-1).”
  • priority (4 rows). Across 4 models with both prices, the median priority price is 1.75× standard. Amazon Bedrock row note: “Priority tier pricing table. Shown for region US East (N. Virginia) on the pricing page (values filled from AWS's published price list for us-east-1).”
  • fast (4 rows). Across 3 models with both prices, the median fast price is 6× standard (range 1.25× standard to 6× standard). Amazon Bedrock row note: “Ultrafast tier, short context, in-region us-east-1 row (six times Standard).”

Cached input: 34 of 108 standard prices include a cache-read price, a median 10% of the input price. What cached input means.

What we recorded about Amazon Bedrock's pricing page: us-east-1 (US East, N. Virginia) on-demand prices. Many tables on aws.amazon.com/bedrock/pricing are filled client-side from AWS's own price list (b0.p.awsstatic.com meteredUnitMaps, region "US East (N. Virginia)"); those values were rendered for us-east-1. Where Bedrock offers both in-region/geo and Global cross-region prices, the in-region / US-geo price is stored and the Global price is given in notes. GPT-5.x/6.x and Grok prices come from the Bedrock model cards linked from the pricing page. Priority/Flex rows are stored only where the page publishes per-model numbers (not where it only says "75% premium" / "50% discount"). Amazon models (Nova, Titan, rerank) are left out (covered elsewhere). Gated models (Claude Mythos, GPT-5.6 Cyber/Daybreak) are left out.

Is Amazon Bedrock cheaper than the model makers' own APIs?

Mostly no. Of 34 standard prices for models whose maker sells its own API, 0 are below the maker's price, 28 above and 6 the same, on the 3:1 blend. The median difference is 10% above the maker. The biggest markup is Command R, 186% above Cohere.

BroadNet Inference and Amazon Bedrock

BroadRouter is BroadNet's own site, so BroadNet is never ranked on this page. BroadNet Inference also serves 2 models that Amazon Bedrock prices: BroadNet Gemma 4 VL 26B at USD 0.10 / USD 0.22 against Amazon Bedrock's USD 0.13 / USD 0.40 (BroadNet is cheaper on the 3:1 blend); GLM 5.3 at USD 0.78 / USD 3.46 against Amazon Bedrock's USD 1.848 / USD 5.808 (BroadNet is cheaper on the 3:1 blend). BroadNet prices are from its rate card, synced 2026-11-14. BroadNet Inference pricing.

More on Amazon Bedrock

Questions

How much does the Amazon Bedrock API cost per token?

Amazon Bedrock's standard prices run from USD 0.04 input and USD 0.04 output per 1M tokens (Voxtral Mini 1.0) to USD 15 / USD 75 (Claude Opus 4.1). The median across 108 standard prices is USD 0.72 input and USD 2.30 output per 1M, which is USD 0.00000072 per input token. Input, cached input and output are billed separately.

What is the cheapest model on Amazon Bedrock?

Voxtral Mini 1.0, at USD 0.04 input and USD 0.04 output per 1M tokens (USD 0.04 on a 3:1 input-to-output blend). The cheapest from other makers: Claude 3 Haiku (Anthropic) at USD 0.25 / USD 1.25; Llama 3.2 1B Instruct (Meta) at USD 0.10 / USD 0.10; GPT OSS Safeguard 20B (OpenAI) at USD 0.07 / USD 0.20.

Is Amazon Bedrock cheaper than other providers?

It depends on the model. Of 68 models that other providers also price, Amazon Bedrock is the cheapest standard price for 7 (1 of them tied with another provider), for example GPT OSS Safeguard 20B, Llama 3.2 11B Vision Instruct and Mistral Large 3; it ties the lowest price, without beating it, for 4 more; another provider is cheaper for 57; and it is the dearest for 35. Its median rank is 4 of a median 5 providers, on a 3:1 input-to-output blend.

Does Amazon Bedrock have batch, cached or priority pricing?

Cached input: yes, for 34 of 108 standard prices, at a median 10% of the input price. Other tiers: batch on 26 rows (median input 50% below standard); flex / spot on 4 rows (median input 50% below standard); priority on 4 rows (median input 1.75× standard); fast on 4 rows (median input 6× standard).

How much does a chatbot cost per month on Amazon Bedrock?

For 2,000 requests a day, 1,500 input tokens (60% cached), 300 output, over 30 days: USD 4.32 on Voxtral Mini 1.0, USD 95.58 on Qwen3 VL 235B A22B (the middle of Amazon Bedrock's list) and USD 2,700 on Claude 3 Opus. Try your own numbers.

Read as MarkdownAsk ChatGPTAsk ClaudeAsk Perplexity