claude-code - 💡(How to fix) Fix [DOCS] Third-party provider docs missing 429 troubleshooting guidance for Bedrock, Vertex, and Foundry [1 participants]

Documentation Type

Missing documentation (feature not documented)

Documentation Location

https://code.claude.com/docs/en/third-party-integrations

Section/Topic

Cloud provider deployment and troubleshooting guidance for 429 rate-limit errors on Amazon Bedrock, Google Vertex AI, and Microsoft Foundry

Current Documentation

The docs currently say:

"Organizations can deploy Claude Code through Anthropic directly or through a cloud provider. This page helps you choose the right configuration."

The Vertex AI guide has a short 429 section:

"If you encounter 429 errors:

• For regional endpoints, ensure the primary model and small/fast model are supported in your selected region
• Consider switching to CLOUD_ML_REGION=global for better availability"

But the Bedrock and Foundry guides' troubleshooting sections do not currently explain how to interpret 429 rate-limit errors for third-party providers or note that status.claude.com only reflects Anthropic-operated services.

What's Wrong or Missing?

Changelog v2.1.111 says Claude Code fixed 429 rate-limit errors on Bedrock, Vertex, and Foundry that were referencing status.claude.com, even though that status page only covers Anthropic-operated providers.

The documentation still does not explain this provider split clearly:

A. Missing cross-provider troubleshooting guidance

The main third-party provider overview does not tell users that 429s on Bedrock, Vertex, and Foundry usually come from provider-specific quotas, capacity, or regional availability rather than Anthropic service health.

B. Inconsistent provider coverage

Vertex AI has partial 429 guidance, but Bedrock and Foundry troubleshooting sections do not have equivalent 429/quota guidance.

C. No explicit status page scope note

The docs do not currently say that status.claude.com is relevant for Anthropic-operated providers, not cloud-provider-hosted deployments.

Suggested Improvement

Add a shared troubleshooting subsection to third-party-integrations for cloud-provider 429 errors, then link to it from the Bedrock, Vertex AI, and Foundry guides.

Suggested content:

429 errors on cloud providers

If you use Amazon Bedrock, Google Vertex AI, or Microsoft Foundry, 429 errors usually come from your cloud provider's quotas, throttling, deployment capacity, or region/model availability.

status.claude.com covers Anthropic-operated services only and should not be used as the primary health check for third-party provider deployments.

Then add provider-specific follow-up bullets:

• Bedrock: check AWS Bedrock quotas/token burndown, model access, and region availability.
• Vertex AI: check Google Cloud quotas and regional/global endpoint support.
• Foundry: check Azure AI Foundry quota/capacity and deployment configuration.

Impact

Medium - Makes feature difficult to understand

Additional Context

Affected Pages:

Total scope: 4 pages affected

Source: Changelog v2.1.111

Exact changelog entry: "Fixed 429 rate-limit errors on Bedrock/Vertex/Foundry referencing status.claude.com (it only covers Anthropic-operated providers)"