Skip to main content
Watch: I Built an AI App Without Leaving Databricks

Cloud platforms - 47 entries

Amazon Bedrock quotas and rate limits

AWS's official docs, read on October 17, 2026, set quotas for Amazon Bedrock. For Anthropic Claude Opus 4.8, the docs list 30,000,000 Cross-region TPM [adjustable], 30,000,000 Global cross-region TPM [adjustable] and 20,000,000 Mantle input TPM [adjustable].

Last verified from docs.aws.amazon.com. What the docs leave out

Sources
2
Tables
34
Entries
47
Verified
Oct 17, 2026

Limits by model, endpoint and plan

Values are copied exactly as the docs show them, units included, in the docs' own order. A dash means the docs list no value for that row.

All 34 tables on this page
  1. bedrock-mantle endpoint (OpenAI Responses API, OpenAI Chat Completions API, Anthropic Messages API)
  2. Anthropic Claude Opus 5, Anthropic Claude Opus 5.5
  3. Anthropic Claude Opus 4.8 and Anthropic Claude Opus 4.7
  4. Anthropic Claude Sonnet 5
  5. Anthropic Claude Sonnet 5.5 and Anthropic Claude Haiku 5.5
  6. Anthropic Claude Fable 5.1 and Anthropic Claude Fable 5
  7. Anthropic Claude Sonnet 4.6 and Anthropic Claude Opus 4.6 V1
  8. Anthropic Claude Opus 4.5
  9. Anthropic Claude Haiku 4.5, Anthropic Claude Sonnet 4.5 V1
  10. Anthropic Claude Sonnet 4.5 V1 1M Context Length
  11. Anthropic Claude 3.7 Sonnet V1
  12. Anthropic Claude 3.5 Haiku
  13. Anthropic Claude 3.5 Sonnet V2
  14. Anthropic Claude 3.5 Sonnet
  15. Anthropic Claude 3 Haiku
  16. Anthropic Claude 3 Sonnet
  17. GPT-5.6 Luna, GPT-5.6 Terra, GPT-5.6 Sol and GPT-6 Astra
  18. GPT-6 Luna and GPT-6 Sol, GPT-6.1 Sol
  19. GPT-5.4 and GPT-5.5
  20. Grok 4.6, Grok 4.7
  21. Moonshot AI Kimi K3
  22. Z.ai GLM 5.3
  23. Amazon Nova 2 Lite, Amazon Nova 2 Omni and Amazon Nova 2 Pro Preview
  24. Amazon Nova Premier V1
  25. Amazon Nova Pro (and Amazon Nova Pro V1 latency-optimized)
  26. Amazon Nova Lite and Amazon Nova Micro
  27. DeepSeek R1 V1 and Meta Llama 4 Maverick V1, Meta Llama 4 Scout V1
  28. Meta Llama 3.3 70B Instruct
  29. Meta Llama 3.1 405B Instruct and Meta Llama 3.1 70B Instruct
  30. Meta Llama 3.1 8B Instruct
  31. Meta Llama 3.2 1B Instruct, Meta Llama 3.2 3B Instruct
  32. Mistral Pixtral Large 25.02 V1 and Writer AI Palmyra X4 V1, Writer AI Palmyra X5 V1
  33. Cohere Embed V4
  34. Twelve Labs Marengo, Twelve Labs Pegasus, TwelveLabs Marengo Embed 3.0

bedrock-mantle endpoint (OpenAI Responses API, OpenAI Chat Completions API, Anthropic Messages API)

bedrock-mantle endpoint (OpenAI Responses API, OpenAI Chat Completions API, Anthropic Messages API) limits
MetricLimit
Quota typesBedrock Mantle input tokens per minute for ${model}; Bedrock Mantle output tokens per minute for ${model}
RPMNot enforced: "The bedrock-mantle endpoint does not enforce requests-per-minute (RPM) quotas."
Input TPM admission checkinput tokens in the request plus max_tokens (or the model-specific maximum if max_tokens is not set); if over quota the request is throttled with an HTTP 429 response
Output TPMcounted during generation; if reached, generation stops and the response is returned with a finish reason indicating the cutoff
  • Separate allocation from bedrock-runtime even for the same model. Cached input tokens read through prompt caching do not count against input TPM. "The endpoint may apply additional internal rate limiting that is not exposed in Service Quotas." Burndown does not apply to models available exclusively on bedrock-mantle. Per-account values are in the Service Quotas console (search bedrock-mantle endpoint).

Anthropic Claude Opus 5, Anthropic Claude Opus 5.5

Model pages: Claude Opus 5, Claude Opus 5.5

Anthropic Claude Opus 5, Anthropic Claude Opus 5.5 limits
MetricLimit
Cross-region TPM [adjustable]30,000,000
Global cross-region TPM [adjustable]30,000,000
On-demand TPM15,000,000
Mantle input TPM [adjustable]20,000,000
Mantle output TPM [adjustable]2,000,000

Anthropic Claude Opus 4.8 and Anthropic Claude Opus 4.7

2 scopes

Anthropic Claude Opus 4.8 and Anthropic Claude Opus 4.7 limits
Model or scopeCross-region TPM [adjustable]
Anthropic Claude Opus 4.830,000,000
Global cross-region TPM [adjustable]
30,000,000
Mantle input TPM [adjustable]
20,000,000
Mantle output TPM [adjustable]
2,000,000
Anthropic Claude Opus 4.710,000,000
Global cross-region TPM [adjustable]
10,000,000
Mantle input TPM [adjustable]
20,000,000
Mantle output TPM [adjustable]
2,000,000
  • Output burndown 15x on bedrock-runtime. No RPM quota on bedrock-runtime.

Anthropic Claude Sonnet 5

Model page: Claude Sonnet 5

Anthropic Claude Sonnet 5 limits
MetricLimit
Cross-region TPM [adjustable]6,000,000
Global cross-region TPM [adjustable]6,000,000
On-demand TPM3,000,000
Mantle input TPM [adjustable]3,000,000
Mantle output TPM [adjustable]300,000

Anthropic Claude Sonnet 5.5 and Anthropic Claude Haiku 5.5

2 scopes

Anthropic Claude Sonnet 5.5 and Anthropic Claude Haiku 5.5 limits
Model or scopeCross-region TPM [adjustable]
Anthropic Claude Sonnet 5.56,000,000
Global cross-region TPM [adjustable]
6,000,000
Anthropic Claude Haiku 5.510,000,000
Global cross-region TPM [adjustable]
10,000,000
  • Burndown rate not named for this model on the token-counting page.

Anthropic Claude Fable 5.1 and Anthropic Claude Fable 5

2 scopes

Anthropic Claude Fable 5.1 and Anthropic Claude Fable 5 limits
Model or scopeCross-region TPM [adjustable]
Anthropic Claude Fable 5.110,000,000
Global cross-region TPM [adjustable]
10,000,000
Mantle input TPM [adjustable]
5,000,000
Mantle output TPM [adjustable]
500,000
Anthropic Claude Fable 54,000,000
Global cross-region TPM [adjustable]
4,000,000
Mantle input TPM [adjustable]
2,000,000
Mantle output TPM [adjustable]
200,000
  • Anthropic Claude Fable 5: Burndown rate not named for this model on the token-counting page.

Anthropic Claude Sonnet 4.6 and Anthropic Claude Opus 4.6 V1

2 scopes

Anthropic Claude Sonnet 4.6 and Anthropic Claude Opus 4.6 V1 limits
Model or scopeCross-region RPM [adjustable]
Anthropic Claude Sonnet 4.610,000
Cross-region TPM [adjustable]
6,000,000
Global cross-region RPM [adjustable]
10,000
Global cross-region TPM [adjustable]
6,000,000
On-demand RPM
5,000
On-demand TPM
3,000,000
Anthropic Claude Opus 4.6 V110,000
Cross-region TPM [adjustable]
3,000,000
Global cross-region RPM [adjustable]
10,000
Global cross-region TPM [adjustable]
3,000,000
On-demand RPM
5,000
On-demand TPM
1,500,000

Anthropic Claude Opus 4.5

Model page: Claude Opus 4.5

Anthropic Claude Opus 4.5 limits
MetricLimit
Cross-region RPM [adjustable]10,000
Cross-region TPM [adjustable]2,000,000
Global cross-region RPM [adjustable]10,000
Global cross-region TPD2,880,000,000
Global cross-region TPM [adjustable]2,000,000
Max TPD (doubled for cross-region calls)1,440,000,000

Anthropic Claude Haiku 4.5, Anthropic Claude Sonnet 4.5 V1

Model page: Claude Haiku 4.5

Anthropic Claude Haiku 4.5, Anthropic Claude Sonnet 4.5 V1 limits
MetricLimit
Cross-region RPM [adjustable]10,000
Cross-region TPM [adjustable]5,000,000
Global cross-region RPM [adjustable]10,000
Global cross-region TPM [adjustable]5,000,000

Anthropic Claude Sonnet 4.5 V1 1M Context Length

Anthropic Claude Sonnet 4.5 V1 1M Context Length limits
MetricLimit
Cross-region RPM [adjustable]1,000
Cross-region TPM [adjustable]1,000,000
Global cross-region RPM [adjustable]1,000
Global cross-region TPD1,440,000,000
Global cross-region TPM [adjustable]1,000,000
Max TPD (doubled for cross-region calls)720,000,000

Anthropic Claude 3.7 Sonnet V1

Anthropic Claude 3.7 Sonnet V1 limits
MetricLimit
Cross-region RPM50
Cross-region TPM [adjustable]50,000
Max TPD (doubled for cross-region calls)720,000,000
On-demand RPM125
On-demand TPM500,000

Anthropic Claude 3.5 Haiku

Anthropic Claude 3.5 Haiku limits
MetricLimit
Cross-region RPM2,000
Cross-region TPM [adjustable]4,000,000
Max TPD (doubled for cross-region calls)2,880,000,000
On-demand latency-optimized RPM100
On-demand latency-optimized TPM500,000
On-demand RPMus-west-1: 400 ; Each of the other supported Regions: 1,000
On-demand TPMus-west-1: 300,000 ; Each of the other supported Regions: 2,000,000

Anthropic Claude 3.5 Sonnet V2

Anthropic Claude 3.5 Sonnet V2 limits
MetricLimit
Cross-region RPM100
Cross-region TPM [adjustable]800,000
Max TPD (doubled for cross-region calls)576,000,000
On-demand RPM50
On-demand TPM400,000

Anthropic Claude 3.5 Sonnet

Anthropic Claude 3.5 Sonnet limits
MetricLimit
Cross-region RPMap-southeast-1: 40 ; Each of the other supported Regions: 100
Cross-region TPM [adjustable]ap-southeast-1: 400,000 ; Each of the other supported Regions: 800,000
On-demand RPMap-southeast-1: 20 ; Each of the other supported Regions: 50
On-demand TPMap-southeast-1: 200,000 ; Each of the other supported Regions: 400,000

Anthropic Claude 3 Haiku

Anthropic Claude 3 Haiku limits
MetricLimit
Cross-region RPMap-southeast-1: 400 ; Each of the other supported Regions: 800
Cross-region TPM [adjustable]ap-southeast-1: 400,000 ; Each of the other supported Regions: 600,000
Max TPD (doubled for cross-region calls)ap-southeast-1: 288,000,000 ; Each of the other supported Regions: 432,000,000
On-demand RPMap-southeast-1: 200 ; Each of the other supported Regions: 400
On-demand TPMap-southeast-1: 200,000 ; Each of the other supported Regions: 300,000

Anthropic Claude 3 Sonnet

Anthropic Claude 3 Sonnet limits
MetricLimit
Cross-region RPM200
Cross-region TPM [adjustable]400,000
On-demand RPM100
On-demand TPM200,000

GPT-5.6 Luna, GPT-5.6 Terra, GPT-5.6 Sol and GPT-6 Astra

3 scopes

GPT-5.6 Luna, GPT-5.6 Terra, GPT-5.6 Sol and GPT-6 Astra limits
Model or scopeCross-region TPM [adjustable]
GPT-5.6 Luna, GPT-5.6 Terra20,000,000
Global cross-region TPM [adjustable]
20,000,000
Mantle input TPM [adjustable]
20,000,000
Mantle output TPM [adjustable]
2,000,000
GPT-5.6 Sol10,000,000
Global cross-region TPM [adjustable]
10,000,000
Mantle input TPM [adjustable]
10,000,000
Mantle output TPM [adjustable]
1,000,000
GPT-6 Astra4,000,000
Global cross-region TPM [adjustable]
4,000,000
Mantle input TPM [adjustable]
1,000,000
Mantle output TPM [adjustable]
100,000
  • GPT-6 Astra: Not named in burndown list (all other models 1:1).

GPT-6 Luna and GPT-6 Sol, GPT-6.1 Sol

2 scopes

GPT-6 Luna and GPT-6 Sol, GPT-6.1 Sol limits
Model or scopeCross-region TPM [adjustable]
GPT-6 Luna4,000,000
Global cross-region TPD
5,760,000,000
Global cross-region TPM [adjustable]
4,000,000
Max TPD (doubled for cross-region calls)
2,880,000,000
Mantle input TPM [adjustable]
2,000,000
Mantle output TPM [adjustable]
200,000
GPT-6 Sol, GPT-6.1 Sol2,000,000
Global cross-region TPD
2,880,000,000
Global cross-region TPM [adjustable]
2,000,000
Max TPD (doubled for cross-region calls)
1,440,000,000
Mantle input TPM [adjustable]
1,000,000
Mantle output TPM [adjustable]
100,000
  • Not named in burndown list (all other models 1:1).

GPT-5.4 and GPT-5.5

2 scopes

GPT-5.4 and GPT-5.5 limits
Model or scopeMantle input TPM [adjustable]
GPT-5.420,000,000
Mantle output TPM [adjustable]
2,000,000
GPT-5.510,000,000
Mantle output TPM [adjustable]
1,000,000
  • Bedrock-mantle quotas only; burndown does not apply.

Grok 4.6, Grok 4.7

Model pages: Grok 4.6, Grok 4.7

Grok 4.6, Grok 4.7 limits
MetricLimit
Cross-region TPM [adjustable]10,000,000
Global cross-region TPM [adjustable]10,000,000

Moonshot AI Kimi K3

Moonshot AI Kimi K3 limits
MetricLimit
Cross-region TPM [adjustable]10,000,000
Global cross-region TPD14,400,000,000
Global cross-region TPM [adjustable]10,000,000
Max TPD (doubled for cross-region calls)14,400,000,000

Z.ai GLM 5.3

Model page: GLM-5.3

Z.ai GLM 5.3 limits
MetricLimit
Cross-region TPM [adjustable]10,000,000
Global cross-region TPM [adjustable]10,000,000
Mantle input TPM [adjustable] (published as "GLM 5.3")10,000,000
Mantle output TPM [adjustable] (published as "GLM 5.3")1,000,000

Amazon Nova 2 Lite, Amazon Nova 2 Omni and Amazon Nova 2 Pro Preview

2 scopes

Amazon Nova 2 Lite, Amazon Nova 2 Omni and Amazon Nova 2 Pro Preview limits
Model or scopeCross-region RPM
Amazon Nova 2 Lite, Amazon Nova 2 Omni2,000
Cross-region TPM [adjustable]
8,000,000
Global cross-region RPM [adjustable]
2,000
Global cross-region TPD
11,520,000,000
Global cross-region TPM [adjustable]
8,000,000
Max TPD (doubled for cross-region calls)
5,760,000,000
Amazon Nova 2 Pro Preview100
Cross-region TPM [adjustable]
1,000,000
Global cross-region RPM [adjustable]
100
Global cross-region TPD
1,440,000,000
Global cross-region TPM [adjustable]
1,000,000
Max TPD (doubled for cross-region calls)
720,000,000

Amazon Nova Premier V1

Amazon Nova Premier V1 limits
MetricLimit
Cross-region RPM500
Cross-region TPM [adjustable]2,000,000
Max TPD (doubled for cross-region calls)1,440,000,000

Amazon Nova Pro (and Amazon Nova Pro V1 latency-optimized)

Model page: Amazon Nova Pro

Amazon Nova Pro (and Amazon Nova Pro V1 latency-optimized) limits
MetricLimit
Cross-region RPM500
Cross-region TPM [adjustable]2,000,000
Max TPD (doubled for cross-region calls)1,440,000,000
On-demand RPM250
On-demand TPM1,000,000
Nova Pro V1: Model invocation max latency-optimized tokens per day57,600,000
Nova Pro V1: On-Demand, latency-optimized RPM10
Nova Pro V1: On-Demand, latency-optimized TPM40,000
  • Output burndown 1:1. Latency-optimized rows are published under the name "Amazon Nova Pro V1".

Amazon Nova Lite and Amazon Nova Micro

2 scopes

Amazon Nova Lite and Amazon Nova Micro limits
Model or scopeCross-region RPM
Amazon Nova Liteus-east-1: 4,000 ; us-east-2: 4,000 ; us-west-1: 4,000 ; us-west-2: 4,000 ; Each of the other supported Regions: 400
Cross-region TPM [adjustable]
us-east-1: 8,000,000 ; us-east-2: 8,000,000 ; us-west-1: 8,000,000 ; us-west-2: 8,000,000 ; Each of the other supported Regions: 400,000
Max TPD (doubled for cross-region calls)
5,760,000,000
On-demand RPM
us-east-1: 2,000 ; eu-west-2: 2,000 ; Each of the other supported Regions: 200
On-demand TPM
us-east-1: 4,000,000 ; eu-west-2: 4,000,000 ; Each of the other supported Regions: 200,000
Amazon Nova Microus-east-1: 4,000 ; us-east-2: 4,000 ; us-west-2: 4,000 ; Each of the other supported Regions: 400
Cross-region TPM [adjustable]
us-east-1: 8,000,000 ; us-east-2: 8,000,000 ; us-west-2: 8,000,000 ; Each of the other supported Regions: 400,000
Max TPD (doubled for cross-region calls)
5,760,000,000
On-demand RPM
us-east-1: 2,000 ; eu-west-2: 2,000 ; Each of the other supported Regions: 200
On-demand TPM
us-east-1: 4,000,000 ; eu-west-2: 4,000,000 ; Each of the other supported Regions: 200,000

DeepSeek R1 V1 and Meta Llama 4 Maverick V1, Meta Llama 4 Scout V1

2 scopes

DeepSeek R1 V1 and Meta Llama 4 Maverick V1, Meta Llama 4 Scout V1 limits
Model or scopeCross-region RPM
DeepSeek R1 V1200
Cross-region TPM [adjustable]
200,000
Meta Llama 4 Maverick V1, Meta Llama 4 Scout V1800
Cross-region TPM [adjustable]
600,000

Meta Llama 3.3 70B Instruct

Meta Llama 3.3 70B Instruct limits
MetricLimit
Cross-region RPM800
Cross-region TPM [adjustable]600,000
On-demand RPM400
On-demand TPM300,000

Meta Llama 3.1 405B Instruct and Meta Llama 3.1 70B Instruct

2 scopes

Meta Llama 3.1 405B Instruct and Meta Llama 3.1 70B Instruct limits
Model or scopeCross-region RPM
Meta Llama 3.1 405B Instruct400
Cross-region TPM [adjustable]
800,000
On-Demand, latency-optimized RPM
100
On-Demand, latency-optimized TPM
40,000
On-demand RPM
200
On-demand TPM
400,000
Meta Llama 3.1 70B Instruct800
Cross-region TPM [adjustable]
600,000
On-Demand, latency-optimized RPM
100
On-Demand, latency-optimized TPM
40,000
On-demand RPM
400
On-demand TPM
300,000

Meta Llama 3.1 8B Instruct

Meta Llama 3.1 8B Instruct limits
MetricLimit
Cross-region RPM1,600
Cross-region TPM [adjustable]600,000
On-demand RPM800
On-demand TPM300,000

Meta Llama 3.2 1B Instruct, Meta Llama 3.2 3B Instruct

Meta Llama 3.2 1B Instruct, Meta Llama 3.2 3B Instruct limits
MetricLimit
Cross-region RPM1,600
Cross-region TPM [adjustable]600,000
Max TPD (doubled for cross-region calls)432,000,000
On-demand RPM800
On-demand TPM300,000

Mistral Pixtral Large 25.02 V1 and Writer AI Palmyra X4 V1, Writer AI Palmyra X5 V1

2 scopes

Mistral Pixtral Large 25.02 V1 and Writer AI Palmyra X4 V1, Writer AI Palmyra X5 V1 limits
Model or scopeCross-region RPM
Mistral Pixtral Large 25.02 V110
Cross-region TPM [adjustable]
80,000
Max TPD (doubled for cross-region calls)
57,600,000
Writer AI Palmyra X4 V1, Writer AI Palmyra X5 V110
Cross-region TPM [adjustable]
150,000
Max TPD (doubled for cross-region calls)
108,000,000

Cohere Embed V4

Cohere Embed V4 limits
MetricLimit
Cross-region RPM2,000
Cross-region TPM [adjustable]300,000
Global cross-region RPM [adjustable]2,000
Global cross-region TPD432,000,000
Global cross-region TPM [adjustable]300,000
Max TPD (doubled for cross-region calls)216,000,000
On-demand RPM1,000
On-demand TPM150,000

Twelve Labs Marengo, Twelve Labs Pegasus, TwelveLabs Marengo Embed 3.0

Twelve Labs Marengo, Twelve Labs Pegasus, TwelveLabs Marengo Embed 3.0 limits
MetricLimit
Twelve Labs Marengo: Cross-region RPM200
Twelve Labs Marengo: On-demand RPM100
Twelve Labs Pegasus: Cross-region RPM [adjustable]120
Twelve Labs Pegasus: Global cross-region RPM [adjustable]240
Twelve Labs Pegasus: On-demand RPM [adjustable]60
TwelveLabs Marengo Embed 3.0: Cross-region RPM [adjustable]1,000
TwelveLabs Marengo Embed 3.0: On-demand RPM [adjustable]500

How to raise your limits

  1. bedrock-mantle endpoint (OpenAI Responses API, OpenAI Chat Completions API, Anthropic Messages API)

    bedrock-mantle quotas are visible in Service Quotas but "quota increase requests are not currently processed through the Service Quotas console". Submit the AWS Support limit increase form, select Amazon Bedrock, and specify the endpoint (bedrock-mantle), Region, model, quota name (input TPM or output TPM) and requested value. Approval depends on whether existing usage justifies the increase; include recent CloudWatch or Service Quotas usage.

  2. Anthropic Claude Opus 5, Anthropic Claude Opus 5.5, Anthropic Claude Opus 4.8 and 44 more

    runtime: Service Quotas; mantle: AWS Support limit increase form

What the docs leave out

  • Per-account applied quota values: Bedrock only publishes default values; the quotas that apply to an account (which "might be lower than the default values") are shown only in the logged-in Service Quotas console.
  • Bedrock-mantle per-model defaults are published in the General Reference only for the models listed there; the docs say mantle quota values "vary by model, account, and AWS Region" and direct users to the Service Quotas console.
  • No usage tiers or automatic tier advancement: Bedrock documents no tier ladder. Defaults may change based on "regional factors, account age, usage history, payment history, fraudulent usage", and "Quota increases aren't granted automatically."
  • No concurrency limit is published for on-demand model inference (concurrency quotas exist only for jobs such as batch, evaluation, data automation and knowledge base ingestion; not listed here).
  • Batch inference, custom model, imported model, Provisioned Throughput, Agents, Knowledge Bases, Guardrails and other non-inference control-plane quotas were left out of this per-model inference extraction.
  • Output-token burndown rates for Anthropic Claude Haiku 5.5, Claude Sonnet 5.5 and Claude Fable 5 are not named on the token-counting page.

FAQ

What are the Amazon Bedrock quotas and rate limits?

AWS's official docs, read on October 17, 2026, set quotas for Amazon Bedrock. For Anthropic Claude Opus 4.8, the docs list 30,000,000 Cross-region TPM [adjustable], 30,000,000 Global cross-region TPM [adjustable] and 20,000,000 Mantle input TPM [adjustable].

How do I raise my Amazon Bedrock quotas?

bedrock-mantle quotas are visible in Service Quotas but "quota increase requests are not currently processed through the Service Quotas console". Submit the AWS Support limit increase form, select Amazon Bedrock, and specify the endpoint (bedrock-mantle), Region, model, quota name (input TPM or output TPM) and requested value. Approval depends on whether existing usage justifies the increase; include recent CloudWatch or Service Quotas usage. runtime: Service Quotas; mantle: AWS Support limit increase form

Where do these Amazon Bedrock numbers come from?

Every value is copied verbatim from AWS's official documentation at docs.aws.amazon.com, read on October 17, 2026. This page lists all 47 entries in 34 tables, each linked to the page it came from.

Related reading

Sources

Last verified . Every value is copied verbatim from the official pages below, read on that date.

Newsletter

Know when limits change

One email per week. New model releases, pricing and limit changes, and what they mean for what you build.