Nova 2 Pro suits AWS-native teams reasoning over huge document or video sets inside Bedrock's compliance boundary, not teams chasing the very top of the reasoning leaderboard. It caps each response at 64,000 output tokens and pairs naturally with Bedrock AgentCore for production agent deployment rather than replacing a self-hosted stack.
Amazon Nova 2 Pro is Amazon's flagship Nova 2 foundation model, built around a 1,000,000-token context window, about four times its predecessor's limit. It adds a built-in code interpreter, live web grounding with citations, and native MCP tool support for agentic workflows on AWS Bedrock.
Where it sits
- $3.44/M$ per 1M tokensBlended price (3:1)Lower is better#43 / 65peer median $1.71/Mvendor price, checked by HokAI
- 116 tok/stokens/sOutput speedHigher is better#14 / 40peer median 90 tok/scited: Artificial Analysis
- 70%% solvedSWE-bench VerifiedHigher is better#23 / 29peer median 78%per source, see benchmark scores
- --% correctGPQA DiamondHigher is better-- / 44peer median 88.3%per source, see benchmark scores
Priced around the middle of the 65 GA models with a published price (rank 43), in the bottom third on SWE-bench Verified (rank 23 of 29), and one of 66 whose vendor states it does not train on customer data. Ranked against GA models; this record is not GA.
Ranks are against GA models on HokAI that publish the same figure; ties share a rank.
Provider: Amazon · Family: Nova 2
Context window: 1,000,000 tokens · Max output: 64,000
Input modalities: text, image, pdf, video, tool-calls · Output: text, tool-calls
About Amazon Nova 2 Pro
Amazon Nova 2 Pro is a multimodal foundation model built by Amazon, announced at AWS re:Invent in December 2025 and released into Bedrock preview on December 2, 2025 (model ID amazon.nova-2-pro-preview-20251202-v1:0). It is the flagship of the Nova 2 family, sitting above Nova 2 Lite and alongside the speech-focused Nova 2 Sonic and multimodal-generation Nova 2 Omni. Amazon has not disclosed the exact architecture (dense versus mixture-of-experts) or parameter count, consistent with its practice for prior Nova generations. Nova 2 Pro was designed to handle highly complex, multistep workloads such as multi-document analysis, video reasoning, and software migrations, positioning it as AWS's answer to frontier reasoning models from OpenAI, Anthropic, and Google rather than a budget option.
On independent third-party evaluation, Nova 2 Pro Preview scores 70.0% on SWE-bench Verified, Amazon's stated benchmark for autonomous software engineering capability. Artificial Analysis places its composite Intelligence Index at 14 out of a field of 78 tracked models, below the frontier tier occupied by GPT-5-series and Claude Opus models, but the model is explicitly optimized for cost and speed rather than topping every leaderboard. Amazon has not published standalone GPQA Diamond, AIME, or MMLU-Pro subscores separate from the composite index, and Nova system cards historically report fewer individual benchmark breakdowns than Anthropic or OpenAI system cards.
Nova 2 Pro's headline architectural feature is its 1,000,000-token context window, a roughly 4x increase over the original Nova Pro's 300K limit, paired with a 64,000-token maximum output. This puts it ahead of GPT-5's and Claude's standard context tiers on raw window size, though Amazon has not published a needle-in-haystack recall benchmark at the full 1M depth, so retrieval quality at maximum context is unverified beyond Amazon's own claims.
The model accepts multimodal input, text, images, PDFs, and video, and generates text output. It ships with built-in agentic tooling: a code interpreter that runs and evaluates Python within the same request, web grounding that retrieves live public information with citations, and support for remote MCP (Model Context Protocol) tools, letting it call external services directly. Function calling and structured output are supported through Bedrock's standard tool-use API. As of mid-2026 there is no confirmed native audio input/output on Nova 2 Pro itself, that capability lives in the separate Nova 2 Sonic model.
Nova 2 Pro is available through Amazon Bedrock's standard API and console, with early access also granted to Amazon Nova Forge customers, AWS's service for pre-training custom Nova variants on proprietary data. It integrates natively with Amazon Bedrock AgentCore for production agent deployment, including AgentCore Policy for blocking unauthorized agent actions, and with Amazon CloudWatch for observability. As a proprietary Amazon model, there are no open weights, no Hugging Face listing, and no self-hosting path; access is exclusively through AWS-managed infrastructure and inherits Bedrock's regional availability and IAM-based authentication.
Amazon's safety approach for Nova models runs through Bedrock's operational controls rather than a published responsible-scaling framework: Bedrock Guardrails for content filtering, AgentCore Policy for agent action governance, and standard Amazon content restrictions on CSAM, weapons, and malware generation. Amazon has not disclosed named third-party red-team partners for the Nova 2 family in the way Anthropic names METR or Apollo Research, and no public system card equivalent to a frontier-lab release has been located for Nova 2 Pro specifically as of mid-2026; the closest public documentation is Amazon Science's technical report for the Nova 2 family and AWS's Bedrock model card page.
Nova 2 Pro is best suited to AWS-native enterprise teams already inside the Bedrock ecosystem who need long-context document and video analysis, agentic coding, or multi-step planning without leaving AWS's compliance boundary, and who value price-performance over topping every reasoning leaderboard. Teams chasing the absolute highest benchmark scores on SWE-bench, GPQA, or AIME should look at Claude Opus or GPT-5-series instead, both of which outscore Nova 2 Pro on independently verified reasoning evals. Teams needing self-hosted or open-weight deployment should also look elsewhere, since Nova 2 Pro has no downloadable weights.
Training data cutoff for Nova 2 Pro has not been publicly disclosed by Amazon in a dated system card; Amazon's general practice trains on a mix of licensed data, public web text, and synthetic reasoning traces, and by default Bedrock does not use customer API inputs or outputs to train the underlying foundation models. Enterprise customers can rely on AWS's broad compliance footprint, SOC 2, ISO 27001, and HIPAA-eligibility apply to Bedrock as an AWS service, though Nova-specific compliance disclosures are not separately published from the parent AWS compliance program.
As a preview model, Nova 2 Pro represents Amazon's move from Nova 1.0's cost-first positioning toward a more capable, still price-competitive reasoning tier. The prior generation, Nova Pro (1.0), priced at $0.80 input / $3.20 output per 1M tokens with a 300K context window, remains available on Bedrock for workloads that don't need Nova 2 Pro's expanded context or agentic tooling. Amazon has signaled that general availability and stabilized pricing for the full Nova 2 lineup will follow the December 2025 preview launch, with Nova 2 Pro reaching broader GA access being the next expected milestone.
Pricing
Nova 2 Pro Preview bills $1.25/$10.00 per 1M input/output tokens on Bedrock's standard on-demand tier, per Artificial Analysis; a CloudPrice snapshot instead shows $2.19 input, $17.50 output, and $0.547 cached-read, likely captured at a different date or region. Confirm the live AWS Bedrock console rate before committing production spend, since the model has not reached general availability.
What a real job costs
| Job | Input | Output | Total |
|---|---|---|---|
| Summarise a 20-page PDF | $0.037 | $0.010 | $0.048 |
| Support reply | $0.0025 | $0.0030 | $0.0055 |
| One coding agent run | $0.250 | $0.200 | $0.450 |
Budgets: 20-page PDF = 30k in / 1k out · Support reply = 2k in / 300 out · Coding agent run = 200k in / 20k out. Computed from the vendor's per-token prices at render time; cached-input discounts are not applied.
Key Features
- 1M-Token Context Window: Roughly 4x the original Nova Pro's 300K limit, enabling multi-document and long-video analysis in a single request.
- Built-In Code Interpreter: Runs and evaluates Python code within the same API call, letting the model self-verify generated code before returning output.
- Web Grounding with Citations: Retrieves live public information and attaches citations, reducing outdated or fabricated factual claims.
- Remote MCP Tool Support: Natively calls Model Context Protocol tools, extending the model's reach into external services without custom scaffolding.
- AgentCore Policy Integration: Blocks unauthorized agent actions against defined policy in milliseconds when deployed through Bedrock AgentCore.
Pros
- 1M-token context window, about 4x the prior Nova Pro generation, for long-document and video workloads.
- 70.0% SWE-bench Verified with competitive pricing versus frontier reasoning models.
- Native AgentCore, CloudWatch, and MCP integration cuts custom agent-infrastructure work for AWS-native teams.
Cons
- Artificial Analysis Intelligence Index of 14 trails frontier competitors like GPT-5-series and Claude Opus on composite reasoning.
- No native audio input or output; voice use cases require pairing with Nova 2 Sonic.
- Still in preview with inconsistent third-party pricing reports; GA pricing not yet locked.
Benchmarks
- SWE-bench Verified: 70% cited: Artificial Analysis · 10 Jul 2026 — Real GitHub issues fixed end to end, % solved.
- AA Intelligence Index: 14 cited: Artificial Analysis · 10 Jul 2026 — Composite of 10 evaluations run by Artificial Analysis, 0 to 100.
- Output speed: 116 tok/s cited: Artificial Analysis · 10 Jul 2026 — Median tokens written per second as measured by Artificial Analysis.
A benchmark is an exam, not the job. Scores transfer unevenly between tasks, so weigh the one closest to your workload and read every figure with its source.
Frequently Asked Questions
What are Amazon Nova 2 Pro's pricing plans in 2026?
Amazon Nova 2 Pro's standard Bedrock rate works out to $1.25 for every million input tokens and $10.00 for every million output tokens, according to Artificial Analysis. A separate tracker, CloudPrice, reports a higher $2.19 input and $17.50 output price with $0.547 for cached reads, which likely reflects a different region or a different snapshot date during the preview window. Because pricing has not been finalized for general availability, check the AWS Bedrock console for the current rate before budgeting a production workload.
Does Amazon Nova 2 Pro have a free plan?
Amazon Nova 2 Pro ships without a dedicated free tier: every request bills through Amazon Bedrock's pay-as-you-go pricing from the first token onward. AWS sometimes issues account-level free credits to new Bedrock customers, but that is a platform-wide promotion, not a Nova-specific plan, and it is not guaranteed to apply. Budget for the per-token rate from day one if you plan to evaluate the model.
What are the best alternatives to Amazon Nova 2 Pro?
GPT-5 undercuts Nova 2 Pro on price at roughly $0.625 input and $5 output per 1M tokens, though its 272,000-token context window is far smaller and it scored 74.9% on SWE-bench Verified at launch. Gemini 3.5 Pro targets a still-larger 2,000,000-token context window and Deep Think reasoning, but remains in limited Vertex preview with no published pricing as of mid-2026. Pick GPT-5 for cheaper high-volume tasks, Gemini 3.5 Pro if you need more context than Nova 2 Pro offers, or stick with Nova 2 Pro for built-in code execution and web grounding without extra scaffolding.
Amazon Nova 2 Pro or Claude Opus 4.8: which should you pick?
Claude Opus 4.8 scores meaningfully higher on SWE-bench Verified, 88.6% versus Nova 2 Pro's more modest result, and is generally the stronger pick for the hardest reasoning and coding tasks. Nova 2 Pro answers back on context parity, since both ship a 1M-token window, and undercuts Claude Opus 4.8 on Bedrock's standard per-token rate. Choose Claude Opus 4.8 if benchmark-topping accuracy matters most; choose Nova 2 Pro if you want AWS-native integration at a lower price for similar context length.
How do you get started with Amazon Nova 2 Pro?
Getting started takes as long as enabling Bedrock model access for amazon.nova-2-pro-preview-20251202-v1:0 in the AWS console, available at launch in the us-east-1 and us-west-2 regions. From there, calls run through Bedrock's standard API using the Python, TypeScript, Java, or Go SDK, the same interface used for other Bedrock models. Teams already inside Bedrock AgentCore can attach Nova 2 Pro to an existing agent pipeline without new infrastructure, though expect occasional API or pricing changes before the model reaches general availability.
Top Alternatives
- GPT-5: Pick GPT-5 if you want the cheaper Bedrock alternative and can live with roughly a third of Nova 2 Pro's context window.
- Claude Opus 4.8: Pick Claude Opus 4.8 if you need the strongest measured reasoning accuracy and can absorb its steeper per-token price.
- Gemini 3.5 Pro: Pick Gemini 3.5 Pro if you need an even larger context window and can tolerate its limited preview status without published pricing.