Amazon Nova Pro API Pricing: High-Capacity Multimodal Bedrock Workhorse
Comprehensive Amazon Nova Pro API pricing analysis ($0.80/M input, $3.20/M output), AWS Bedrock provisioned throughput, multimodal document extraction, and enterprise SLAs.
Full specs, context window and API limits →How much does Amazon Nova Pro cost per million tokens?
Amazon Nova Pro costs $0.80 per million input tokens and $3.20 per million output tokens ($1.40/M blended at 3:1). Amazon Bedrock premier multimodal model delivering balanced speed, deep document comprehension, and seamless AWS cloud integration. Verified 2026-09-08.
How much does Amazon Nova Pro cost per 1,000 requests?
Computed from generated token pricing. Each row assumes the listed input and output tokens per request; this model has no measured verbosity factor, so the unadjusted output estimate is shown.
| Request shape | Input tokens | Output tokens | Cost / 1,000 requests |
|---|---|---|---|
| Short | 100 | 50 | $0.2400 |
| Medium | 1,000 | 500 | $2.4000 |
| Long | 4,000 | 2,000 | $9.6000 |
Formula: ((input price × input tokens) + (output price × output tokens × verbosity factor)) ÷ 1,000,000 × 1,000. Assumptions: short 100/50, medium 1,000/500, long 4,000/2,000 input/output tokens per request. Unadjusted — no measured verbosity factor is available.
Three model-specific pricing decisions
Nova Pro is evaluated using only unit-compatible text shapes; multimodal conversion is unavailable when the registry does not publish a comparable unit.
1. Pro / Lite / Micro fixed text and multimodal-shaped crossover
| Workload | Pro / 100K | Lite / 100K | Micro / 100K | Unit boundary |
|---|---|---|---|---|
| Text chat | $304.00 | $22.80 | $13.30 | Token-compatible |
| Image-shaped request | $304.00 | $22.80 | $13.30 | Image unit unavailable; text proxy only |
| Long context | $1920.00 | $144.00 | $84.00 | Context-tier price unavailable |
2. Region, on-demand, batch, cache, and provisioned-throughput matrix
| Mechanic | AWS/Bedrock evidence | Decision |
|---|---|---|
| Region | Model row is USD; region surcharge unavailable | Do not generalize across regions |
| On-demand | Dated base rate | Use only listed unit |
| Batch / cache | Unavailable | No discount default |
| Provisioned throughput | Unavailable | Capacity price not inferred |
3. Quality/retry-adjusted Pro uplift boundary over Lite
Required uplift by workload
| Fixed workload | Amazon Nova Pro | Amazon Nova Lite | Boundary |
|---|---|---|---|
| Chat | $304.00 | $22.80 | 10% accepted-result uplift needed to justify premium |
| Image-shaped proxy | $304.00 | $22.80 | 20% accepted-result uplift needed to justify premium |
| Long context | $1920.00 | $144.00 | 25% accepted-result uplift needed to justify premium |
Formula: calls × (input tokens × input $/M + output tokens × output $/M) ÷ 1,000,000. The uplift threshold is a decision input, not a measured quality claim.
Verified 2026-06-14. Luna is the data owner for this rendered decision module. “Unavailable” means the current dated registry has no model-specific evidence; it is not a zero. First-party price source · Run this scenario in the playground.
All results are server-rendered for Amazon Nova Pro; formulas expose fixed inputs and missing evidence remains visibly unavailable.
Batch 62 · exact-model pricing decision contributions · verified 2026-09-07
Exact model boundary: Amazon Amazon Nova Pro (nova-pro). Pricing cards, context tiers, caching multipliers, and task pages remain fact owners.
AWS Bedrock on-demand vs provisioned throughput break-even
Frozen Batch 62 scenario board. Formula / deterministic rule: on_demand = calls * ((in * 0.80 + out * 3.20) / 1M); provisioned = model_units * hourly_rate * 730 Boundary: Owns Bedrock on-demand versus provisioned throughput financial break-evens.
| Frozen scenario / field ID | Exact identity and evidence fields | Result | State |
|---|---|---|---|
batch62-nova-pro-m1-r1low-traffic unpredictable workload | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=low-traffic unpredictable workload; monthly call volume; on-demand cost; provisioned cost; utilization break-even point; recommended purchase; financial verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — low-traffic unpredictable workload is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
batch62-nova-pro-m1-r2100K consistent queries/month | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=100K consistent queries/month; monthly call volume; on-demand cost; provisioned cost; utilization break-even point; recommended purchase; financial verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — 100K consistent queries/month is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
batch62-nova-pro-m1-r31M steady-state enterprise volume | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=1M steady-state enterprise volume; monthly call volume; on-demand cost; provisioned cost; utilization break-even point; recommended purchase; financial verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — 1M steady-state enterprise volume is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
batch62-nova-pro-m1-r410M high-concurrency peak load | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=10M high-concurrency peak load; monthly call volume; on-demand cost; provisioned cost; utilization break-even point; recommended purchase; financial verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — 10M high-concurrency peak load is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
batch62-nova-pro-m1-r5uncommitted provisioned hour idle | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=uncommitted provisioned hour idle; monthly call volume; on-demand cost; provisioned cost; utilization break-even point; recommended purchase; financial verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — uncommitted provisioned hour idle is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
batch62-nova-pro-m1-r6unsupported provisioned model unit tier | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=unsupported provisioned model unit tier; monthly call volume; on-demand cost; provisioned cost; utilization break-even point; recommended purchase; financial verdict; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — unsupported provisioned model unit tier has no matched, dated bilateral observation. | FAIL CLOSED — manual, probe, or source evidence required |
First-party provenance: AWS Bedrock Nova pricing; verification date 2026-09-07. Missing or conflicting joins fail closed.
Multimodal vision and document OCR ingestion on Bedrock
Frozen Batch 62 scenario board. Formula / deterministic rule: ocr_cost = pages * ((tokens_per_page * 0.80 + extracted_json * 3.20) / 1M) Boundary: Owns AWS Bedrock multimodal document extraction economics on Nova Pro.
| Frozen scenario / field ID | Exact identity and evidence fields | Result | State |
|---|---|---|---|
batch62-nova-pro-m2-r11K invoice structured extractions | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=1K invoice structured extractions; document type; page volume; input tokens; completion tokens; total spend; cost per page; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — 1K invoice structured extractions is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
batch62-nova-pro-m2-r210K medical record scans | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=10K medical record scans; document type; page volume; input tokens; completion tokens; total spend; cost per page; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — 10K medical record scans is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
batch62-nova-pro-m2-r350K technical engineering diagrams | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=50K technical engineering diagrams; document type; page volume; input tokens; completion tokens; total spend; cost per page; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — 50K technical engineering diagrams is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
batch62-nova-pro-m2-r4200K loan application forms | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=200K loan application forms; document type; page volume; input tokens; completion tokens; total spend; cost per page; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — 200K loan application forms is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
batch62-nova-pro-m2-r5resolution downsampling penalty | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=resolution downsampling penalty; document type; page volume; input tokens; completion tokens; total spend; cost per page; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — resolution downsampling penalty is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
batch62-nova-pro-m2-r6unsupported image MIME type | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=unsupported image MIME type; document type; page volume; input tokens; completion tokens; total spend; cost per page; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — unsupported image MIME type has no matched, dated bilateral observation. | FAIL CLOSED — manual, probe, or source evidence required |
First-party provenance: Amazon Nova documentation; verification date 2026-09-07. Missing or conflicting joins fail closed.
Nova Pro vs Nova Lite enterprise routing threshold
Frozen Batch 62 scenario board. Formula / deterministic rule: blended_spend = pro_pct * pro_cost + lite_pct * lite_cost; Lite costs $0.06/M in, $0.24/M out Boundary: Owns Bedrock routing decisions between Nova Pro and Nova Lite.
| Frozen scenario / field ID | Exact identity and evidence fields | Result | State |
|---|---|---|---|
batch62-nova-pro-m3-r1simple text summarization (Lite optimal) | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=simple text summarization (Lite optimal); task category; Nova Lite spend; Nova Pro spend; cost ratio; accuracy requirement; routing recommendation; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — simple text summarization (Lite optimal) is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
batch62-nova-pro-m3-r2customer FAQ matching (Lite optimal) | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=customer FAQ matching (Lite optimal); task category; Nova Lite spend; Nova Pro spend; cost ratio; accuracy requirement; routing recommendation; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — customer FAQ matching (Lite optimal) is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
batch62-nova-pro-m3-r3complex contract analysis (Pro optimal) | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=complex contract analysis (Pro optimal); task category; Nova Lite spend; Nova Pro spend; cost ratio; accuracy requirement; routing recommendation; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — complex contract analysis (Pro optimal) is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
batch62-nova-pro-m3-r4multimodal visual inspection (Pro optimal) | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=multimodal visual inspection (Pro optimal); task category; Nova Lite spend; Nova Pro spend; cost ratio; accuracy requirement; routing recommendation; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — multimodal visual inspection (Pro optimal) is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
batch62-nova-pro-m3-r580% Lite / 20% Pro hybrid router | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=80% Lite / 20% Pro hybrid router; task category; Nova Lite spend; Nova Pro spend; cost ratio; accuracy requirement; routing recommendation; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — 80% Lite / 20% Pro hybrid router is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
batch62-nova-pro-m3-r6untested routing classifier accuracy | model=nova-pro; provider=Amazon; slug=nova-pro; scenario=untested routing classifier accuracy; task category; Nova Lite spend; Nova Pro spend; cost ratio; accuracy requirement; routing recommendation; prompt/config/input/output/result/cache/checkpoint/artifact hashes=required; evidence=2026-09-07; measurement versus assumption=explicit | Unavailable — untested routing classifier accuracy is a frozen fixture pending exact identity, configuration, and denominator joins. | UNTESTED — assumption cannot establish a verdict |
First-party provenance: AWS Bedrock Nova pricing; verification date 2026-09-07. Missing or conflicting joins fail closed.
Method and limitations: formulas are deterministic; observed and assumed inputs are labeled; no missing provider, host, account, region, realm, alias, snapshot, revision, weight, artifact, control, tool, modality, workload, rate-period, timestamp, or result is transferred. Run the Amazon Nova Pro Batch 62 scenario →
nova-proAmazon Nova Pro API Pricing: High-Capacity Multimodal Bedrock Workhorse
Amazon Nova Pro costs $0.80 per million input tokens and $3.20 per million output tokens ($1.40/M blended at 3:1). Amazon Bedrock premier multimodal model delivering balanced speed, deep document comprehension, and seamless AWS cloud integration. Verified 2026-09-08.
Amazon Nova Pro delivers deep multimodal reasoning at a competitive $1.40/M blended rate.
| Scenario | Rendered Evidence & Bounds |
|---|---|
| Scenario 1 | Enterprise contract layout analysis (16K in, 2K out): $0.019200 per document |
| Scenario 2 | Complex PDF invoice table extraction (8K in, 1K out): $0.009600 per invoice |
| Scenario 3 | Multimodal chart & diagram QA (4K in, 600 out): $0.005120 per asset review |
| Scenario 4 | Customer support dialogue turn (2K in, 400 out): $0.002880 per message exchange |
| Scenario 5 | Full-text corporate policy search synthesis (32K in, 3K out): $0.035200 per dossier |
| Scenario 6 | Monthly 50M token enterprise multimodal tier: $70.00 infrastructure spend |
On-demand billing provides superior cost agility for enterprise workloads under 8B tokens/mo.
| Scenario | Rendered Evidence & Bounds |
|---|---|
| Scenario 1 | On-demand spend at 100M tokens/mo: $140.00 total monthly billing |
| Scenario 2 | On-demand spend at 1B tokens/mo: $1,400.00 total monthly billing |
| Scenario 3 | 1 Model Unit 1-month commitment ($13,100/mo): requires >9.4 billion tokens/mo to break even |
| Scenario 4 | Provisioned throughput eliminates all rate limit throttles during critical corporate peak hours |
| Scenario 5 | Guarantees sub-second latency SLAs for real-time customer-facing banking applications |
| Scenario 6 | Recommended strategy: on-demand billing for all variable traffic fleets below 8B tokens/mo |
Native AWS PrivateLink and IAM integration eliminate data egress risks and security overhead.
| Scenario | Rendered Evidence & Bounds |
|---|---|
| Scenario 1 | Data never traverses the public internet: communicates entirely over AWS PrivateLink |
| Scenario 2 | Native integration with AWS IAM role-based access control eliminates API key rotation risk |
| Scenario 3 | FedRAMP High, HIPAA, and SOC 2 Type II certified out-of-the-box on AWS Bedrock |
| Scenario 4 | Eliminates cross-cloud data egress fees: saves up to $0.09/GB compared to external SaaS APIs |
| Scenario 5 | Automated CloudWatch audit logging tracks 100% of LLM queries for forensic compliance |
| Scenario 6 | Optimal infrastructure foundation for defense, healthcare, and financial services |
How fast is Amazon Nova Pro?
How much does Amazon Nova Pro cost at scale?
| Tokens / month | Est. cost (blended 3:1) |
|---|---|
| 100,000 | $0.14 |
| 1,000,000 | $1.40 |
| 10,000,000 | $14.00 |
| 100,000,000 | $140.00 |
How does Amazon Nova Pro compare with other models?
What is Amazon Nova Pro best for?
What should you explore next for Amazon Nova Pro?
Which Amazon Nova Pro head-to-head comparisons are available?
What are common questions about Amazon Nova Pro?
Is Amazon Nova Pro cheaper than Gemini 3.7 Flash?
Amazon Nova Pro costs $1.40/M blended tokens, Gemini 3.7 Flash costs $1.50/M — Amazon Nova Pro is cheaper.
How much does 1 million tokens cost with Amazon Nova Pro?
At a 3:1 input:output ratio, 1 million blended tokens costs approximately $1.40. Pure input costs $0.80/M; pure output costs $3.20/M.
What does Amazon Nova Pro cost at high volume?
At 100 million blended tokens a month, Amazon Nova Pro costs approximately $140.00. See the cost-at-scale table below for other volumes.