diff --git a/build_with_arch/agent.html b/build_with_arch/agent.html index 97289f9d..9656fe41 100755 --- a/build_with_arch/agent.html +++ b/build_with_arch/agent.html @@ -417,7 +417,7 @@ the user’s intent.

diff --git a/build_with_arch/multi_turn.html b/build_with_arch/multi_turn.html index c0513a2c..f3011c13 100755 --- a/build_with_arch/multi_turn.html +++ b/build_with_arch/multi_turn.html @@ -353,7 +353,7 @@ that you can test and modify locally for multi-turn RAG scenarios.

diff --git a/build_with_arch/rag.html b/build_with_arch/rag.html index 6e88c8ff..4b8fd1d4 100755 --- a/build_with_arch/rag.html +++ b/build_with_arch/rag.html @@ -301,7 +301,7 @@ minutes. For additional details on how to build multi-turn RAG applications plea diff --git a/concepts/llm_providers/client_libraries.html b/concepts/llm_providers/client_libraries.html index d821fc30..a9a6f6c8 100755 --- a/concepts/llm_providers/client_libraries.html +++ b/concepts/llm_providers/client_libraries.html @@ -583,7 +583,7 @@ Implement fallback logic for better reliability:

diff --git a/concepts/llm_providers/llm_providers.html b/concepts/llm_providers/llm_providers.html index 39d9d476..ded43978 100755 --- a/concepts/llm_providers/llm_providers.html +++ b/concepts/llm_providers/llm_providers.html @@ -300,7 +300,7 @@ Use your preferred client library without changing existing code (see
-

© 2025, Katanemo Labs, Inc Last updated: Oct 21, 2025. 

+

© 2025, Katanemo Labs, Inc Last updated: Oct 22, 2025. 

diff --git a/concepts/llm_providers/model_aliases.html b/concepts/llm_providers/model_aliases.html index d9ee8f60..cc1f98a0 100755 --- a/concepts/llm_providers/model_aliases.html +++ b/concepts/llm_providers/model_aliases.html @@ -434,7 +434,7 @@ diff --git a/concepts/llm_providers/supported_providers.html b/concepts/llm_providers/supported_providers.html index 92fa1323..4a3254f2 100755 --- a/concepts/llm_providers/supported_providers.html +++ b/concepts/llm_providers/supported_providers.html @@ -718,6 +718,36 @@ Any provider that implements the OpenAI API interface can be configured using cu +
+

Amazon Bedrock

+

Provider Prefix: amazon_bedrock/

+
+
API Endpoint: Arch automatically constructs the endpoint as:
    +

  • Non-streaming: /model/{model-id}/converse

    • +

  • Streaming: /model/{model-id}/converse-stream

    • +
+
+
+

Authentication: AWS Bearer Token + Base URL - Get your API Keys from AWS Bedrock Console → Discover → API Keys.

+

Supported Chat Models: All Amazon Bedrock foundation models including Claude (Anthropic), Nova (Amazon), Llama (Meta), Mistral AI, and Cohere Command models.

+
llm_providers:
+  # Amazon Nova models
+  - model: amazon_bedrock/us.amazon.nova-premier-v1:0
+    access_key: $AWS_BEARER_TOKEN_BEDROCK
+    base_url: https://bedrock-runtime.us-west-2.amazonaws.com
+    default: true
+
+  - model: amazon_bedrock/us.amazon.nova-pro-v1:0
+    access_key: $AWS_BEARER_TOKEN_BEDROCK
+    base_url: https://bedrock-runtime.us-west-2.amazonaws.com
+
+  # Claude on Bedrock
+  - model: amazon_bedrock/us.anthropic.claude-3-5-sonnet-20241022-v2:0
+    access_key: $AWS_BEARER_TOKEN_BEDROCK
+    base_url: https://bedrock-runtime.us-west-2.amazonaws.com
+
+
+

Qwen (Alibaba)

Provider Prefix: qwen/

@@ -733,7 +763,7 @@ Any provider that implements the OpenAI API interface can be configured using cu # Multiple deployments - model: qwen/qwen3-coder access_key: $DASHSCOPE_API_KEY - base_url: "https://dashscope-intl.aliyuncs.com", + base_url: "https://dashscope-intl.aliyuncs.com"
@@ -883,6 +913,7 @@ Any provider that implements the OpenAI API interface can be configured using cu
  • Providers Requiring Base URL