Inference proxy returns upstream quota/auth errors for all models

[zixxtech]

Your issue will be reviewed by a maintainer and labeled for further action.
To complete your request, follow the bot-actions to report this on the support E-mail box and follow the outlined steps. Alternatively, you can report this issue on the official web page to an agent for assistance.
@testedmedia

---

             Issue Description 

Setup: Provisioned automaton with API key, $25 credits loaded, sandbox running (8db0343db2efbb9b367dee7bbd929592). Wallet funded with USDC on Base.

Problem: All inference calls fail with upstream provider errors. Every model returns a different auth/quota issue:

OpenAI models (gpt-4.1, gpt-4.1-mini, gpt-4.1-nano, gpt-5-mini, gpt-5-nano, o3-mini, o4-mini):

429: "You exceeded your current quota, please check your plan and billing details."

Anthropic models (claude-haiku-4.5, claude-sonnet-4.5):

"Your credit balance is too low to access the Anthropic API. Please go to Plans & Billing to upgrade or purchase credits."

Moonshot (kimi-k2.5):

"Invalid Authentication"

Gemini (gemini-2.5-flash, gemini-2.5-pro, gemini-3-flash, gemini-3-pro):

Listed as available: false in /v1/models

Result: Automaton boots, detects $25 credits, calls inference, gets 429/503 on every model, hits 5 consecutive errors, enters sleep state. No model works.

Endpoint tested: https://inference.conway.tech/v1/chat/completions

Request format:

{
"model": "gpt-4.1",
"messages": [{"role": "user", "content": "Say hello"}],
"max_tokens": 10
}
Auth: API key in Authorization header (confirmed working — /v1/credits/balance, /v1/sandboxes, /v1/models all respond correctly).

This appears to be a platform-wide issue with upstream API key quotas/auth, not account-specific. Related to #17 (billing was also broken, now fixed).

[github-actions]

For quick review and assistance, contact the hub agent on the live chat page to report your issue via Web Support page. Alternatively send an E-mail to report this via Gitactivemailbox@gmail.com to ensure total review with the technical support.

Team🚀