Amazon · Amazon Nova · Small
Amazon Nova Lite pricing and limits
A very low-cost multimodal model on Bedrock, with Nova Micro below it for text only. Both suit high-volume workloads inside AWS.
| Line | Tokens | Rate / 1M | Cost |
|---|---|---|---|
| Input, uncached | 2,000 | $0.060 | $0.00012 |
| Input, cache read | 2,000 | $0.015 | $0.00003 |
| Output | 300 | $0.24 | $0.00007 |
- Input / 1M
- $0.060
- Output / 1M
- $0.24
- Cached / 1M
- $0.015
- Cache write / 1M
- n/a
- Blended 3:1
- $0.10
- Context
- 300K
What drives Amazon Nova Lite cost
At this price, volume and prompt length drive spend. Bedrock prompt caching discounts repeated prefixes, and batch inference halves the rate for asynchronous jobs.
Nova Micro is text-only and cheaper still; routing pure text calls to Micro and keeping Lite for image or video input is a simple way to lower blended cost.
Cost of a typical call
Take a extraction call with 4,000 input tokens, 2,000 of them served from cache, and 300 output tokens. At Amazon Nova Lite list prices ($0.060 input, $0.015 cached input, $0.24 output per 1M) one call costs $0.00022, against $0.00031 with no cache hits. At 3,000,000 calls a month that is $666, or $936 uncached.
Blended at a 3:1 input to output ratio, Amazon Nova Lite costs $0.10 per 1M tokens. Routing the same call to the fallback in the policy below, Amazon Nova Micro, costs $0.00013, which is what a tenant pays you in COGS once they cross the hard cap and PALM SOUQ downgrades instead of blocking.
Cost comparison
Amazon Nova Lite against three alternatives
The same extraction call priced on Amazon Nova Micro, Gemini 2.5 Flash-Lite, GPT-5 nano. Move the sliders to match your traffic; models without a cache-read rate bill every input token in full.
Benchmark quality per task before switching. A cheaper model that needs more steps or longer outputs can cost more per finished task than its rate suggests, which is why PALM SOUQ reports cost per feature, not only per token.
Amazon Nova Lite this guide
$0.00022 per call
$555 / mo
baseline
Amazon Nova Micro
$0.00013 per call
$324 / mo
-42% vs Amazon Nova Lite
Gemini 2.5 Flash-Lite
$0.00037 per call
$925 / mo
+67% vs Amazon Nova Lite
GPT-5 nano
$0.00023 per call
$575 / mo
+4% vs Amazon Nova Lite
List prices from the PALM SOUQ LLM Price Index, computed per call as uncached input + cache reads + output. Reasoning tokens count as output on reasoning models.
Rate limits by usage tier
Per-account, per-region Bedrock quotas for requests and tokens per minute, adjustable via Service Quotas. Smaller models usually have higher default quotas.
Exact limits vary by account tier and change over time, so we do not reprint them here. Check the limits page for your own account in the Amazon documentation at docs.aws.amazon.com/bedrock. PALM SOUQ enforces your own per-tenant limits in front of the provider's, so a single customer cannot consume the shared allowance that every other tenant depends on.
Circuit-breaker best practices
Use request quotas per tenant to protect shared AWS account quotas, and a token velocity ceiling to catch loops quickly.
Track Lite and Micro as separate cost lines using the X-PalmSouq-Feature header so you can see which features justify multimodal pricing.
When a breaker trips, the gateway answers in under 50 ms with HTTP 429 and X-PalmSouq-Reason: circuit_breaker_triggered, before the request reaches Amazon. Your agent framework sees a normal rate-limit response and stops, and the audit log records which heuristic fired and for which tenant.
# palmsouq.policy.yaml
policy: nova-lite-pro
match:
model: nova-lite
tier: pro # from X-PalmSouq-Tier
budgets:
monthly_usd: 40 # per tenant
soft_cap_pct: 80 # webhook to your app
on_hard_cap: fallback # or: block (HTTP 429)
fallback:
- nova-micro
breakers:
tool_thrashing: { identical_calls: 3, window_steps: 8 }
token_velocity: { max_tokens_per_s: 50000, window_s: 10 }
semantic_echo: { repeats: 3 }
max_steps: 8
cache:
meter_reads: true- budgetsPer-tenant monthly dollar cap with an 80% soft-cap webhook, keyed by X-PalmSouq-Tenant.
- fallbackAt the hard cap, requests reroute to Amazon Nova Micro ($0.061 blended per 1M) instead of failing.
- breakersAll four loop heuristics on, evaluated in the request path with under 5 ms added latency.
Where it fits
Classification, captioning, extraction and routing at very high volume within AWS.
Provider documentation: docs.aws.amazon.com/bedrock
Amazon Nova Lite pricing questions
How much does Amazon Nova Lite cost per million tokens?
At list price Amazon Nova Lite costs $0.060 per 1M input tokens and $0.24 per 1M output tokens, with cached input at $0.015 per 1M. Prices were last reviewed in the PALM SOUQ LLM Price Index and are for the standard tier and shortest context band.
Does Amazon Nova Lite support prompt caching?
Yes. Cached input is billed at $0.015 per 1M tokens against $0.060 for uncached input. PALM SOUQ records cached tokens per tenant from the provider's usage data, so cache savings show up in each customer's COGS.
What are the Amazon Nova Lite rate limits?
Per-account, per-region Bedrock quotas for requests and tokens per minute, adjustable via Service Quotas. Smaller models usually have higher default quotas. For your account's exact numbers, see the Amazon documentation at docs.aws.amazon.com/bedrock.
How do I stop runaway Amazon Nova Lite agent loops?
Use request quotas per tenant to protect shared AWS account quotas, and a token velocity ceiling to catch loops quickly. The sample policy on this page enables all four PALM SOUQ heuristics with a max step cap of 8 and falls back to Amazon Nova Micro at the hard cap.
Is there a batch discount for Amazon Nova Lite?
50% off on-demand rates for supported models through Bedrock batch inference. Batch traffic suits evaluations, backfills and nightly enrichment jobs that do not need a real-time answer.
Govern Amazon Nova Lite spend per customer.
Point your Amazon client at gateway.palmsouqltd.com, pass X-PalmSouq-Tenant, and every Amazon Nova Lite token is attributed, capped and protected by circuit breakers.
14-day free production trial · 100k requests included · no card required