What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Amazon Nova 2 Lite is the best-documented of the three models and the one to build against first on Amazon Bedrock. AWS publishes a current Bedrock model card, a developer guide, and a model ID for it, and it is called through the Converse and InvokeModel APIs. Nova 2 Pro and Nova 2 Omni have only AWS’s December 2, 2025 preview announcements. Those describe what AWS intended each model to do, but they do not establish whether either model is generally available, where, or through which API today. The status statements below reflect AWS’s published material as it stood when this guide was written, so confirm them on the live pages before committing to a design.
The three models at a glance
The table separates what AWS documents as a current technical reference from what it said about Pro and Omni at launch. Where a value does not appear in that material, the cell says “Not stated” and names the source.
As an Amazon Associate I earn from qualifying purchases.
| Attribute | Nova 2 Lite | Nova 2 Pro | Nova 2 Omni |
|---|---|---|---|
| Status in AWS material | Active on its Bedrock model card; launch date December 2, 2025 | Preview at announcement (December 2, 2025); current status not stated | Preview at announcement (December 2, 2025); current status not stated |
| Documentation available | Bedrock model card, developer guide, AI service card, launch announcement | December 2025 announcement only | December 2025 announcement only |
| Positioning | Everyday multimodal work: document processing, customer support, automation | Highly complex, multistep tasks | All-in-one multimodal reasoning and image generation |
| Input modalities | Text, image, video | Not stated (December 2025 announcement) | Text, image, video, speech |
| Output modalities | Text | Not stated (December 2025 announcement) | Text, image |
| Context window | 1M tokens (Bedrock model card) | Not stated (December 2025 announcement) | 1M tokens (December 2025 announcement) |
| Maximum output | 64K tokens (Bedrock model card) | Not stated (December 2025 announcement) | Not stated (December 2025 announcement) |
| Model ID | amazon.nova-2-lite-v1:0 |
Not stated (December 2025 announcement) | Not stated (December 2025 announcement) |
| Documented API access | Converse and InvokeModel via bedrock-runtime |
Not stated (December 2025 announcement) | Not stated (December 2025 announcement) |
| Reasoning control | Extended thinking, off by default; low, medium, or high | Not stated (December 2025 announcement) | Not stated (December 2025 announcement) |
| Early access at launch | Not applicable; generally listed as active | Nova Forge customers; others contact their AWS account team | Nova Forge customers could request early access |
Nova 2 Lite: the documented entry point
Lite is the model to evaluate first for a new Bedrock integration. Its listing is active, and AWS publishes what a developer needs to call it: the model ID, the supported APIs, context and output limits, and a feature matrix for its Bedrock endpoint.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Where Lite is positioned
AWS describes Lite for simple automation, document processing, and customer support. Its service card adds multi-turn interactions, retrieval-augmented generation, and agentic workflows. The December 2025 announcement also named customer service chatbots, business process automation, software engineering, and business intelligence as possible uses. These are AWS’s descriptions of intended use. They are not independent performance findings, so fit has to be judged with your own evaluation.
#1 Best Overall
- Your favorite music and content – Play music, audiobooks, and podcasts from Amazon Music, Apple Music, Spotify and others or via Bluetooth throughout your home.
- Alexa is happy to help – Ask Alexa for weather updates and to set hands-free timers, get answers to your questions and even hear jokes. Need a few extra minutes in the morning? Just tap your Echo Dot to snooze your alarm.
- Keep your home comfortable – Control compatible smart home devices with your voice and routines triggered by built-in motion or indoor temperature sensors. Create routines to automatically turn on lights when you walk into a room, or start a fan if the inside temperature goes above your comfort zone.
- Do more with device pairing – Fill your home with music using compatible Echo devices in different rooms, or create a home theatre system with Fire TV.
- Say goodbye to drop-offs and buffering - With eero Built-in, Echo Dot doubles as a mesh wifi extender, adding up to 1,000 sq. ft. of wifi coverage to your existing eero network.
A lifecycle date to plan around
The Bedrock model card lists Lite’s end-of-life as “no sooner than” December 2, 2026. That is an earliest possible date, not a scheduled retirement, and AWS’s material cited here does not give a retirement date. Because the earliest date is roughly two months away from the time this article was written, a production design should include a migration path rather than treating Lite as a long-term default.
Calling Lite on Amazon Bedrock
AWS documents Lite access through the Converse and InvokeModel operations and recommends the bedrock-runtime endpoint for new applications where possible. Converse provides one message format for conversational calls; InvokeModel is the lower-level operation, whose request body follows the model’s native format as described in the developer guide.
Prerequisites
- An AWS account with Amazon Bedrock access and access to the model in the Region you use.
- IAM permissions to invoke the model. Converse calls require
bedrock:InvokeModel, and ConverseStream calls requirebedrock:InvokeModelWithResponseStream. - An installed SDK. The example below uses Boto3 for Python.
Model IDs and inference geography
| Model ID | Where AWS shows it | Use it when |
|---|---|---|
amazon.nova-2-lite-v1:0 |
Bedrock model card, as the published model ID | Referencing the base model ID as listed by AWS |
us.amazon.nova-2-lite-v1:0 |
Developer guide, as the example for a U.S. Region | Calling from a U.S. Region with US cross-Region inference |
global.amazon.nova-2-lite-v1:0 |
Boto3 Converse example in AWS’s Lite announcement (sample Region us-east-1) |
Global cross-Region inference, as in AWS’s sample |
The developer guide also describes geography-prefixed inference IDs for use outside the U.S. This article does not list those prefixes, so copy the exact ID for your Region from AWS’s current model listing rather than hard-coding one from an example.
Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallRank #2
- Your favorite music and content – Play music, audiobooks, and podcasts from Amazon Music, Apple Music, Spotify and others or via Bluetooth throughout your home.
- Alexa is happy to help – Ask Alexa for weather updates and to set hands-free timers, get answers to your questions and even hear jokes. Need a few extra minutes in the morning? Just tap your Echo Dot to snooze your alarm.
- Keep your home comfortable – Control compatible smart home devices with your voice and routines triggered by built-in motion or indoor temperature sensors. Create routines to automatically turn on lights when you walk into a room, or start a fan if the inside temperature goes above your comfort zone.
- Do more with device pairing – Fill your home with music using compatible Echo devices in different rooms, or create a home theatre system with Fire TV.
- Say goodbye to drop-offs and buffering - With eero Built-in, Echo Dot doubles as a mesh wifi extender, adding up to 1,000 sq. ft. of wifi coverage to your existing eero network.
A minimal Converse request
- Create a
bedrock-runtimeclient for your Region. - Call
conversewith the model ID that matches your inference geography. - Set
maxTokensininferenceConfigto bound output length and cost. - Read the text blocks from the response. If reasoning content is present, it will not be a text block, so loop over the content blocks and keep those that contain text instead of reading index 0.
import boto3nnclient = boto3.client('bedrock-runtime', region_name='us-east-1')nnresponse = client.converse(n modelId='global.amazon.nova-2-lite-v1:0', # use 'us.amazon.nova-2-lite-v1:0' for US cross-Region inferencen messages=[n {n 'role': 'user',n 'content': [{'text': 'Summarize the refund policy below in three bullet points: ...'}],n }n ],n inferenceConfig={'maxTokens': 512, 'temperature': 0.3},n)nnfor block in response['output']['message']['content']:n if 'text' in block:n print(block['text'])nFor streamed output, use converse_stream with the same arguments. The model card lists response streaming as supported for Lite on bedrock-runtime.
Lite feature support on bedrock-runtime
Feature support is endpoint-specific and can change, so treat the table below as a starting checklist. Verify each feature you depend on for the Region and endpoint you plan to use.
| Feature | Status for Lite | Where AWS states it |
|---|---|---|
| Response streaming | Supported | Bedrock model card |
| Guardrails | Supported | Bedrock model card |
| Prompt optimization | Supported | Bedrock model card |
| Client-side tool calling | Supported | Bedrock model card |
| Structured outputs | Listed as unsupported | Bedrock model card |
| Agents | Listed as unsupported | Bedrock model card |
| Amazon Nova Grounding (built-in web access) | Described as built in; confirm for your endpoint | AI service card |
| Code Interpreter (Python calculations) | Described as built in; confirm for your endpoint | AI service card |
Extended thinking: off by default, redacted, and billed
Lite’s extended thinking is disabled unless you enable it. When enabled, you choose a reasoning effort of low, medium, or high, and the mode is set in the request as the Nova 2 developer guide describes. Two constraints shape how you use it:
Rank #3
- Meet Echo Dot Max: Experience rich room-filling sound that automatically adapts to your space and fine-tunes playback. Features a built-in smart home hub and Omnisense technology for highly personalized experiences.
- Music to your ears: With nearly 3x the bass versus Echo Dot (2022 release), it fits beautifully in any space, delivering your personal sound stage with deep bass and enhanced clarity. Listen to streaming services, such as Amazon Music, Apple Music, Spotify, and SiriusXM. Encore!
- Do more with device pairing: Connect compatible Echo smart speakers and smart displays in different rooms, or pair with a second Echo Dot Max to enjoy even richer sound
- Simple smart home control: Set routines, pair and control lights, locks, and thousands of smart home devices that work with Alexa without needing a separate smart home hub. With Omnisense technology, you can activate routines via temperature or presence detection.
- Say goodbye to drop-offs and buffering - With eero Built-in, Echo Dot Max doubles as a mesh wifi extender, adding up to 1,000 sq. ft. of wifi coverage to your existing eero network.
- Reasoning is redacted. AWS says Lite’s internal reasoning is removed from the response, so you cannot read it to audit or debug how an answer was reached.
- Reasoning tokens are billable. They count toward billing, so higher effort raises request cost even when the visible answer is short.
Reserve higher effort for steps where a better answer justifies the extra tokens, and measure the difference on your own inputs rather than assuming it.
Customization paths
AWS’s Nova 2 service card describes supervised fine-tuning and reinforcement fine-tuning on Amazon Bedrock and Amazon SageMaker AI. The Nova product page also describes deploying custom models on SageMaker AI. Three cautions apply:
- Confirm that your model, Region, and fine-tuning method are currently eligible. Service-level restrictions and model availability are set in AWS’s current documentation, not in this guide.
- The AI service card may not apply to customized SageMaker models or to models hosted on nova.amazon.com, so run a separate evaluation for those deployments.
Nova 2 Pro: announced for multistep work
AWS announced Nova 2 Pro in preview on December 2, 2025 for highly complex, multistep tasks. Its examples were multi-document analysis, video reasoning, and software migrations. At announcement, early access was available to Nova Forge customers, and other organizations were told to contact their AWS account team.
Rank #4
- Your favorite music and content – Play music, audiobooks, and podcasts from Amazon Music, Apple Music, Spotify and others or via Bluetooth throughout your home.
- Alexa is happy to help – Ask Alexa for weather updates and to set hands-free timers, get answers to your questions and even hear jokes. Need a few extra minutes in the morning? Just tap your Echo Dot to snooze your alarm.
- Keep your home comfortable – Control compatible smart home devices with your voice and routines triggered by built-in motion or indoor temperature sensors. Create routines to automatically turn on lights when you walk into a room, or start a fan if the inside temperature goes above your comfort zone.
- Do more with device pairing – Fill your home with music using compatible Echo devices in different rooms, or create a home theatre system with Fire TV.
- Say goodbye to drop-offs and buffering - With eero Built-in, Echo Dot doubles as a mesh wifi extender, adding up to 1,000 sq. ft. of wifi coverage to your existing eero network.
Those statements describe the preview as launched. AWS’s material available for this guide does not confirm Pro’s current status, model ID, regional availability, or API support, so do not design around a 2025 access path without checking it. If Pro’s task profile fits your workload, ask your AWS account team for current availability in writing before you plan a build.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Nova 2 Omni: announced as an all-in-one multimodal model
AWS introduced Omni in preview on December 2, 2025 with this description: “Amazon Nova 2 Omni is an all-in-one model for multimodal reasoning and image generation.” The announcement listed:
Free tools Windows power users keep installed
One-click scans. No signup required.
- Inputs: text, image, video, and speech.
- Outputs: text and image.
- Image generation and editing.
- Speech transcription.
- Text processing in more than 200 languages and speech input in 10 languages, both as AWS launch figures.
Of the three models, Omni is the only one AWS describes as generating images. Its current status, interfaces, and regional access are not established by that announcement, so treat image generation as a capability to verify rather than a confirmed shipping feature. At launch, Nova Forge customers could request early access.
Best Value
- MEET ECHO SPOT - A sleek smart alarm clock with Alexa and big vibrant sound. Ready to help you wake up, wind down, and so much more.
- CUSTOMIZABLE SMART CLOCK - See time, weather, and song titles at a glance, control smart home devices, and more. Personalize your display with your favorite clock face and fun colors.
- BIG VIBRANT SOUND - Enjoy rich sound with clear vocals and deep bass. Just ask Alexa to play music, podcasts, and audiobooks. See song titles and touch to control your music.
- EASE INTO THE DAY - Set up an Alexa routine that gently wakes you with music and gradual light. Glance at the time, check reminders, or ask Alexa for weather updates.
- KEEP YOUR HOME COMFORTABLE - Control compatible smart home devices. Just ask Alexa to turn on lights or touch the screen to dim. Create routines that use motion detection to turn down the thermostat as you head out or open the blinds when you walk into a room.
Choosing among Lite, Pro, and Omni
Pro and Omni cannot be compared on a current technical specification today, because AWS has not published a current reference for either. For most teams starting now, Lite is the default, and the decisions below follow from that.
- Text output from text, image, or video input. Lite fits, and it is the option the current documentation supports for a production build.
- Image output or speech input. Lite cannot generate images under its documented output modalities. Evaluate Omni, and confirm its availability and API access before you design around it.
- Highly complex, multistep work. Lite with extended thinking at a higher effort is the documented route to deeper reasoning now, at the cost of redacted, billable reasoning tokens. Pro is the model AWS announced for this profile, but its access must be confirmed first.
- A tuned model. Lite is the model with documented supervised and reinforcement fine-tuning in AWS’s service card. Confirm eligibility for your Region.
Cost, quotas, and vendor claims
Billing and quotas
- On-demand inference for Nova 2 is billed on input and output tokens, with reasoning tokens counted as described above.
- Default quotas vary by model and Region. Increases are requested through Service Quotas.
- Current rates are on the live Amazon Bedrock pricing page. This article gives no per-token price, because AWS’s material available for it does not establish a figure that would remain valid.
- AWS’s Nova 2 material does not include latency figures, so measure latency on your own workload.
Vendor-reported results
AWS’s Nova product page reports up to 60% improvement in tool-calling efficiency for code modernization, based on AWS Transform evaluations of Nova 2 Lite. This is AWS’s own claim for one workload type. It is not a general benchmark and not a comparison with other models.
The same page includes a customer testimonial from Trey Pezzetti, AI Product Manager at PGA TOUR: “With Amazon Nova 2 Lite on Amazon Bedrock, we can cover every shot, every player, every hole, every week at a cost that makes it viable to run continuously and build on top of.” It describes one customer’s experience and is not a cost study.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsThis guide includes no independent benchmark results for any of the three models. Treat AWS’s performance statements as vendor claims until your own tests confirm them.
Before you build
AWS’s AI Service Card frames a Nova 2 deployment as a managed Bedrock service and asks customers to define the business problem, stakeholders, workflow, system inputs and outputs, sources of variation, and the impact of errors before judging a model. Use that list as the outline for your evaluation.
Quick Recap
- Check Lite’s model card status and its end-of-life date.
- Confirm the model ID and inference prefix for your Region, and check AWS’s regional availability page.
- Verify each endpoint feature you depend on, including streaming, Guardrails, tool calling, and any built-in tools.
- Estimate cost with your own token volumes, using live Bedrock pricing and the quotas in Service Quotas.
- For Pro or Omni, obtain current status, model IDs, Region availability, and API support from AWS before planning.
- Test on representative inputs, including the failure cases that matter most to your business.
n
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
