Claude Code Subagents Fail on Vertex AI (Cowork) — Model Routing Hardcoded to Anthropic API Format

What this error means

All subagent types fail on Cowork — model routing hardcoded to claude-haiku-4-5-20251001 (Anthropic API format), rejected before consuming any tokens is a Claude Code failure pattern reported for developers trying to fix claude code subagents failing on vertex ai / cowork deployment with anthropic api format error. Based on the imported evidence, treat this as a tool-specific troubleshooting page rather than a generic API error.

Why this happens

Official Claude Code issue #58274 (2026-05-12) reports all subagent types fail on Cowork/Vertex AI (3P) deployment. Subagents hardcoded to Anthropic API format model claude-haiku-4-5-20251001 are rejected. Main session with Opus 4.6 works fine.

Common causes

• Enterprise users deploying Claude Code via Vertex AI (3P) find all subagent types completely non-functional. The main session works but every subagent attempt fails before token consumption. Critical for enterprise adopters paying for Vertex AI integration.
• Official Claude Code issue #58274 (2026-05-12) reports all subagent types fail on Cowork/Vertex AI (3P) deployment. Subagents hardcoded to Anthropic API format model claude-haiku-4-5-20251001 are rejected. Main session with Opus 4.6 works fine.

Quick fixes

1. Confirm the exact error signature matches All subagent types fail on Cowork — model routing hardcoded to claude-haiku-4-5-20251001 (Anthropic API format), rejected before consuming any tokens.
2. Check the Claude Code account, local tool state, and provider configuration involved in the failing workflow.
3. Check the build output, project root, and deployment platform configuration before redeploying.

Platform/tool-specific checks

• Verify the command, editor, extension, or API client that produced the error.
• Compare local settings with CI, deployment, or editor-level settings when the error appears in only one environment.
• Avoid deleting credentials, local model data, or project settings until the failing scope is clear.

Step-by-step troubleshooting

1. Capture the exact error message and the command, editor action, or request that triggered it.
2. Check whether the failure is account/auth, quota/rate, model/provider, local runtime, or deployment configuration.
3. Review the source evidence below and compare it with your environment.
4. Apply one change at a time and rerun the smallest failing action.
5. Keep the working fix documented for the team or deployment environment.

How to prevent it

• Keep provider/tool configuration documented.
• Record non-secret diagnostics such as tool version, provider name, model name, and command path.
• Add a lightweight check before CI or production workflows depend on the tool.

Sources checked

• https://github.com/anthropics/claude-code/issues/58274

Evidence note: Official Claude Code issue #58274 (2026-05-12) reports all subagent types fail on Cowork/Vertex AI (3P) deployment. Subagents hardcoded to Anthropic API format model claude-haiku-4-5-20251001 are rejected. Main session with Opus 4.6 works fine.

Related errors

• Claude Code Vertex AI subagent configuration
• Claude Code Cowork model routing override

FAQ

What should I check first?

Start with the exact All subagent types fail on Cowork — model routing hardcoded to claude-haiku-4-5-20251001 (Anthropic API format), rejected before consuming any tokens text and the smallest action that reproduces it.

Can I ignore this error?

No. Treat it as a failed Claude Code workflow until the root cause is understood.

Is this guaranteed to have one fix?

No. The imported evidence supports the troubleshooting path above, but tool behavior can vary by account, plan, version, provider, and local configuration.

How do I know the fix worked?

Rerun the same command, editor action, or request. The fix is working when that action completes without All subagent types fail on Cowork — model routing hardcoded to claude-haiku-4-5-20251001 (Anthropic API format), rejected before consuming any tokens.