Subscription-Based Authentication for Venice CLI

Howdy there! I'm a Venice Pro subscriber, and I'd love to use the models my subscription already covers outside the web app, through your official, open-source CLI and in small personal tools of my own, without a second, separately metered inference budget.

The ask: subscription auth for personal use. At minimum, let subscribers authenticate and access their subscription-included chat models through the official CLI. Ideally, a subscription-scoped API token that also works in small personal tools. Either way, the guardrails below would limit abuse and discourage reselling or proxying while keeping usage within reasonable personal limits.

Why I want it:

I primarily write long-form fiction. Chat history lives in the browser (and any encrypted exports), which fits Venice's privacy model but makes multi-device work painful and any kind of continuity borderline inaccessible. With subscription access, I could build a local client with my own cross-device history synchronization and my own needs in mind, keeping conversations under my control rather than requiring Venice to host them.
It enables things the web app can't do: accessibility setups (on-device STT/TTS paired with Venice's text models), long-form drafting tools, and small hobby projects.
It gives technical, privacy-focused people who avoid web apps entirely a reason to subscribe. Your own February 2026 CLI proposal names this "privacy-conscious developer demographic" and cites their demand, scripting, and competitive parity as reasons the CLI exists; this would extend those reasons to subscribers.

Why not just use credits? I know the API is a separate, usage-based product, and that paid plans include monthly credits for agentic chat, premium models, non-text generation, and API calls. At Pro's allotment, though, those credits go fast! After about a month of very careful use, sticking mostly to models my plan already includes, I'm down to 5 of my 100. Meanwhile, those same models are unlimited in the web app's non-agentic chat. Higher tiers raise the ceiling but still meter what the web app treats as unlimited. What I'm asking for is parity with the access I already pay for, not a bigger allowance.

Precedent Anthropic includes Claude Code with consumer subscriptions, while OpenAI goes further by allowing participating third-party apps to use ChatGPT plan allowances. Both demonstrate that subscription-backed inference can coexist with separately metered APIs. I'd love to see Venice offer similar flexibility for its existing CLI and, ideally, small personal integrations. Given Venice's privacy-first positioning, extending it to personal projects seems like a natural next step. (To be clear, I'm not suggesting turning Venice CLI into an agentic coding tool, there's plenty of those already.)

Guardrails:

Revocable, account-scoped credentials; low concurrency and usage limits appropriate for one person's interactive use; eligible models only.
Personal, noncommercial use only, under an explicitly authorized subscription integration policy.
Hitting a subscription limit never silently falls back to paid credits; that would be an explicit opt-in via CLI config.
The metered API stays as-is for higher-volume use and separately billed models.
Eligibility and limits are enforced server-side.

I know the real work here is on Venice's auth and billing side. The CLI already has bearer-token auth infrastructure that could potentially be extended for subscription credentials. If the CLI needs alteration, I'm happy to contribute on GitHub!

Thanks for considering it! I'm enjoying Venice a lot and would like to enjoy it more. :)