Introduce OpenAI-compatible local proxy for OAuth providers
TL;DR
The new `hermes proxy` command creates a local endpoint that translates OpenAI API calls to authenticated OAuth providers like Claude Pro and ChatGPT Pro. This allows tools like Aider and Cline to use existing subscriptions.
What changed
The new hermes proxy command launches a local endpoint that converts standard OpenAI API requests into authenticated calls for OAuth-based services such as Claude Pro and ChatGPT Pro. It launched on 19 May 2026 and targets tools including Aider and Cline that expect an OpenAI-compatible base URL.
Why it matters
Existing paid subscriptions can now power local coding agents without extra API keys or per-token charges from OpenAI. This shifts cost control back to users who already pay monthly for Claude or ChatGPT while keeping their preferred editor workflows intact. It also reduces vendor lock-in for developers who route tasks through multiple front ends.
How to use it
Install or update Hermes Agent via the existing bash script on a VPS running Python 3.10 or later. Run hermes proxy in the terminal, then set the base URL in Aider or Cline to http://localhost:8080/v1 with your existing OAuth session active. No new configuration files are required beyond the standard gateway YAML already used for messaging platforms.
Watch for
Confirm the bet if Aider and Cline releases add native support for the proxy within the next month. The approach breaks if OAuth providers start blocking local endpoints or change token scopes. Expect a follow-up command that adds rate-limit mirroring and session refresh so long-running agent tasks survive token expiry.
Who this matters for
- Vibe Builders: Use your existing Claude Pro sub to power Aider and Cline without paying extra for API tokens.
- Developers: Point your local agents to http://localhost:8080/v1 to route OpenAI calls through OAuth sessions.
Harsh’s take
This is a clever arbitrage play for the local agent ecosystem. By wrapping OAuth sessions in an OpenAI-compatible proxy, Hermes Agent effectively removes the double-tax of paying for a web subscription and an API key simultaneously. It is a direct response to the high token costs associated with long-context coding tasks in tools like Aider.
Operators should watch the stability of these session tokens. OAuth providers have every incentive to rotate keys or block non-browser traffic to protect their API revenue. If this proxy method holds, it turns a standard $20 monthly sub into a high-volume backend for headless agents.
It is a pragmatic win for builders who want to maximize their existing spend before committing to enterprise-scale API bills.
by Harsh Desai
About Hermes Agent
View the full Hermes Agent page →All Hermes Agent updatesGo deeper
More AI news
- Daily RoundupGLM 5.3 and Vercel agent tools land, Gemini SAT practice, plus fresh models on Hugging Face
Google and Vercel released new agent and model features while Hugging Face and Replicate added trending models; OpenAI tightened safety after agent issues.
- Daily RoundupVercel-Origin beta and GPT-5.6 Sol discount, plus agent tools for builders
Vercel rolled out Origin repository deployments and a month-long GPT-5.6 Sol discount on AI Gateway while Product Hunt featured new agent layers and Hugging Face highlighted an uncensored Qwen model, with industry moves around data use and automation exits.
- Weekly DigestHermes Agent v0.20 voice and A2A protocol, OpenClaw mobile updates, and multi-agent tools to test now
Hermes Agent rolled out v0.20 releases with voice streaming, A2A interoperability, desktop artifacts, and self-recovery tools while OpenClaw added Android switching, iOS caching, and plugin provenance checks across the week.