How do I enable models for Amazon Bedrock (gateway) in Atlas?

To enable models for Amazon Bedrock (gateway), you must first enable them in the AWS Bedrock console for your specific account and region. After enabling, run `atlas models amazon-bedrock` to verify what models your Atlas instance can resolve.

What is the largest context window available with Amazon Bedrock (gateway)?

The largest context window available through Amazon Bedrock (gateway) is 3,500,000 tokens, provided by Llama 4 Scout. This is the most extensive context window Atlas can access via any provider.

What are the pricing rates for models on Amazon Bedrock (gateway)?

Pricing for Amazon Bedrock (gateway) models is passthrough at AWS rates. For example, Nova Lite costs $0.06 per Mtok input and $0.24 per Mtok output. Llama 4 Scout is $0.17 per Mtok input and $0.66 per Mtok output, while DeepSeek-R1 is $1.35 per Mtok input and $5.40 per Mtok output.

Does Atlas handle AWS authentication for Bedrock?

Yes, Atlas supports the full AWS authentication chain for Amazon Bedrock (gateway). This includes standard methods like `AWS_PROFILE`, `AWS_ACCESS_KEY_ID`, `AWS_BEARER_TOKEN_BEDROCK`, web identity token files, and container credentials.

Can I switch between different models on Amazon Bedrock (gateway) within Atlas?

Yes, Atlas lets you switch the active model and provider on the fly. Once you have enabled multiple models in the Bedrock console and they are resolved by Atlas, you can select your preferred model from `/models` within the Atlas interface.

What is the benefit of a single IAM boundary for Bedrock models?

The primary benefit of a single IAM boundary for Amazon Bedrock (gateway) models is simplified procurement. A single approval process covers access to a wide range of frontier models including Claude, Llama, DeepSeek, Qwen, and Nova, streamlining administrative overhead.

Atlas with Amazon Bedrock (gateway) in 2026

Atlas with Amazon Bedrock (gateway) provides access to a diverse suite of frontier models including Claude, Llama 4, DeepSeek-R1, Qwen3-Coder, and Amazon's own Nova, all secured behind a single AWS IAM boundary. This integration offers the largest context window Atlas can reach, up to 3.5 million tokens with Llama 4 Scout, at passthrough AWS rates like Nova Lite at $0.06 per Mtok input.

Why choose Amazon Bedrock (gateway) for Atlas?

Amazon Bedrock (gateway) stands out in 2026 by consolidating access to a diverse array of frontier models, including Claude, Llama 4, DeepSeek-R1, Qwen3-Coder, and Nova, all under a single AWS IAM boundary. This simplifies procurement, as one approval covers every model, and offers the largest context window Atlas can reach, up to 3.5 million tokens.

Choosing Amazon Bedrock (gateway) for Atlas means gaining access to a powerful collection of frontier models, including Claude, Llama 4, DeepSeek, Qwen3-Coder, and Amazon's Nova, all managed through a single AWS IAM boundary. This unified access streamlines the procurement process, as a single approval can cover all these diverse models. Furthermore, Amazon Bedrock (gateway) exposes Llama 4 Scout with an exceptional 3,500,000 token context window, which is the largest context window Atlas can reach through any provider. Atlas also supports the full AWS authentication chain, including `AWS_PROFILE`, `AWS_ACCESS_KEY_ID`, `AWS_BEARER_TOKEN_BEDROCK`, web identity token files, and container credentials, ensuring direct integration with existing AWS environments.

What are the context window and pricing for Amazon Bedrock (gateway) models?

Atlas with Amazon Bedrock (gateway) provides a variable context window, reaching up to 3.5 million tokens on Llama 4 Scout, which is the largest window Atlas can access through any provider in 2026. Pricing is passthrough at AWS rates, with Nova Lite at $0.06 per Mtok input and Llama 4 Scout at $0.17 per Mtok input.

The context window for models available via Amazon Bedrock (gateway) varies significantly by model. Notably, Llama 4 Scout is exposed with a 3,500,000 token context on Bedrock, representing the largest window Atlas can reach through any provider. Pricing for Amazon Bedrock (gateway) models is passthrough at standard AWS rates. For instance, Nova Lite is priced at $0.06 per Mtok for input and $0.24 per Mtok for output. Llama 4 Scout costs $0.17 per Mtok for input and $0.66 per Mtok for output. DeepSeek-R1, another powerful option, is priced at $1.35 per Mtok for input and $5.40 per Mtok for output, allowing developers to make informed cost-benefit decisions.

What are the tradeoffs when using Amazon Bedrock (gateway) with Atlas?

While Amazon Bedrock (gateway) offers significant advantages, developers in 2026 should be aware of two key tradeoffs. Models are opt-in per account and per region, meaning `atlas models amazon-bedrock` might show an empty or partial list until enabled in the console. Region prefixes on model IDs, such as 'us.' or 'eu.', can also be a common source of misconfiguration.

When integrating Amazon Bedrock (gateway) with Atlas, developers should consider a couple of specific tradeoffs. First, models within Bedrock are opt-in per AWS account and per region. This means that running `atlas models amazon-bedrock` might initially display an empty or incomplete list of available models until you explicitly enable them in the AWS Bedrock console for your chosen region. This extra setup step is crucial for initial access. Second, region prefixes on model IDs, such as 'us.', 'eu.', or 'global.', are a common source of misconfiguration when switching regions or setting up new environments. While Atlas handles the region prefix on the model ID for you once selected, understanding this underlying mechanism is important for troubleshooting.

When should I consider a different model provider for Atlas?

If immediate, zero-configuration access to models is a priority, or if your workflow in 2026 requires avoiding region-specific model configurations, you might consider an alternative to Amazon Bedrock (gateway). The opt-in nature of Bedrock models per account and region, along with potential misconfiguration from region prefixes, introduces initial setup steps that other providers might bypass.

While Amazon Bedrock (gateway) offers powerful capabilities, there are scenarios where a different model provider for Atlas might be more suitable. If your primary concern is immediate, frictionless access to models without any console-based setup, the opt-in requirement for Bedrock models per account and region might introduce unwanted friction. Similarly, if your development workflow frequently involves switching regions or requires a simpler configuration model that avoids region prefixes on model IDs, an alternative provider might offer a more straightforward experience. For developers prioritizing speed of deployment over the consolidated access and extensive context windows offered by Bedrock, exploring other options could be beneficial.

Setup

01Authenticate as you already do for AWS: `export AWS_PROFILE=my-profile` and `export AWS_REGION=us-east-1`
02Enable the models you want in the Bedrock console (they are opt-in per region).
03Run `atlas models amazon-bedrock` to see what your account actually resolved.
04Select a model from `/models`; Atlas handles the region prefix on the model ID for you.

Frequently asked questions

How do I enable models for Amazon Bedrock (gateway) in Atlas?: To enable models for Amazon Bedrock (gateway), you must first enable them in the AWS Bedrock console for your specific account and region. After enabling, run `atlas models amazon-bedrock` to verify what models your Atlas instance can resolve.
What is the largest context window available with Amazon Bedrock (gateway)?: The largest context window available through Amazon Bedrock (gateway) is 3,500,000 tokens, provided by Llama 4 Scout. This is the most extensive context window Atlas can access via any provider.
What are the pricing rates for models on Amazon Bedrock (gateway)?: Pricing for Amazon Bedrock (gateway) models is passthrough at AWS rates. For example, Nova Lite costs $0.06 per Mtok input and $0.24 per Mtok output. Llama 4 Scout is $0.17 per Mtok input and $0.66 per Mtok output, while DeepSeek-R1 is $1.35 per Mtok input and $5.40 per Mtok output.
Does Atlas handle AWS authentication for Bedrock?: Yes, Atlas supports the full AWS authentication chain for Amazon Bedrock (gateway). This includes standard methods like `AWS_PROFILE`, `AWS_ACCESS_KEY_ID`, `AWS_BEARER_TOKEN_BEDROCK`, web identity token files, and container credentials.
Can I switch between different models on Amazon Bedrock (gateway) within Atlas?: Yes, Atlas lets you switch the active model and provider on the fly. Once you have enabled multiple models in the Bedrock console and they are resolved by Atlas, you can select your preferred model from `/models` within the Atlas interface.
What is the benefit of a single IAM boundary for Bedrock models?: The primary benefit of a single IAM boundary for Amazon Bedrock (gateway) models is simplified procurement. A single approval process covers access to a wide range of frontier models including Claude, Llama, DeepSeek, Qwen, and Nova, streamlining administrative overhead.

Try SeaShell in your terminal

The terminal-native AI coding agent. Free core, single binary.

Install SeaShell

Related guides

Atlas for Pandas: Terminal-Native AI Coding in 2026

Atlas is a terminal-native AI coding agent for Pandas. Vectorize df.apply, fix chained assignment under Copy-on-Write, and pin DataFrames with assert_frame_equal.

Audit a Repo with Parallel Subagents in Atlas (2026 Workflow)

How to audit a repo with parallel subagents in Atlas in 2026: the task tool launches explore subagents in their own sessions, so only conclusions return to your context.

Atlas for Node.js in 2026

Node.js developers in 2026 can adopt Atlas, the terminal-native AI coding agent, for secure, efficient development. Leverage local embeddings, AST indexing, and robust safety features.

Atlas for Scala in 2026

Atlas is a terminal-native AI coding agent for Scala in 2026. Run it in a project with a build.sbt, let it read your traits and implicits, and approve every diff.

Atlas vs Cosine: Choosing Your AI Coding Agent in 2026

Compare Atlas and Cosine AI coding agents for 2026. Atlas offers terminal-native TUI, explicit change review, and BYO model keys. Cosine features its Lumen models and a 30.08% SWE-bench record.

Plan a Multi-File Change Before Editing with Atlas in 2026

How to plan a multi-file change with Atlas in 2026: the plan agent denies all edit tools, you research with codebase_search and lsp, then plan_exit hands off.

Diagnose a Hanging or Long-Running Command with Atlas in 2026

How to diagnose a hanging command with Atlas in 2026: the bash tool races every command against a timeout and tells you whether it is slow or blocked on input.

Atlas for Go in 2026

Atlas, the terminal-native AI coding agent, empowers Go developers in 2026 with intelligent code understanding, safe refactoring, and robust testing capabilities.

Browse this resource hub