LotsDirectoryLotsDirectory
Z

Zelyx

One line of code. Every model.

Listed from the product's public website (zelyx.app)

Is this your product? Claim it to edit the listing, post updates and answer questions. Prove it with an email address at the product's domain (the listing becomes Claimed), or put the LotsDirectory badge on your website (the listing becomes Verified, and its link to your website is followed while the badge is live).

Description

One line of code. Every model. No shock bills. The spend-control layer for AI agents.

About Zelyx

What is Zelyx?

Zelyx is AI cost enforcement infrastructure. It sits in the call path between an application and AI providers, enforces budgets, and kills a runaway loop before the provider is billed.

What Zelyx does

Point an SDK at Zelyx with a Zelyx key and a proxy base URL. Two lines change, with no SDK swap and no config file. Every call is checked in real time. When a daily cap is hit, Zelyx returns HTTP 402 before the provider is called. Alerts fire at 80%.

Who Zelyx is for

Zelyx is for teams running LLM traffic in production who need budgets, kill switches, and cost breakdowns on every request. Design partners in the private beta can start on Free.

How Zelyx works

An admin connects a provider key, which is validated and stored encrypted in a vault. A Zelyx proxy key is shown once and cannot be retrieved again. On each call the proxy verifies the key, decrypts the provider key, runs the gate, forwards the request if approved, streams the response, and records cost, tokens, latency, and model. The app never sees the provider key.

Key features

  • One proxy for every provider and every call. The decision happens in the call path, in milliseconds, before the invoice.
  • Supported providers include OpenAI, Anthropic, Google, Azure AI Foundry, Groq, Together, Mistral, Cohere, DeepSeek, xAI, OpenRouter, Fireworks, Perplexity, or any custom slug.
  • Budgets can be set on company, team, project, model, or per-person keys, and traffic can be tagged with headers.
  • The gate checks budgets, model policies, and payment patterns. A budget that is exceeded returns HTTP 402 before the provider is contacted, so no tokens are consumed.
  • Loop detection and a kill switch stop a runaway loop. Alerts fire at 80%.
  • Provider keys stay in an encrypted vault. The Zelyx key is shown once.
  • Streaming works with the provider SDK. Cost, tokens, latency, and model are recorded in Overview and Explorer.
  • Quickstart is described as under five minutes, and only two lines of code change.

Pricing

As listed on zelyx.app, checked 19 October 2026. Prices can change; see the site for current terms.

  • Zelyx is currently in private beta. Design partners get Individual free for 3 months.
  • The about page says to start on Free.
  • Zelyx charges for the control layer. Provider keys and token costs stay with the team.
  • Reach out for a plan that fits the team.

Facts from zelyx.app, checked 19 October 2026

  • Private beta, with design partners.
  • Individual is free for 3 months for design partners.
  • HTTP 402 is returned before the provider is called when a cap is hit.
  • Alerts fire at 80%.
  • Provider keys are stored encrypted.
  • Continue with Google.
  • Quickstart is described as under five minutes.

Written from Zelyx's own website (Homepage, Features, About, Docs). The maker can claim this listing to edit it.

Categories

Frequently asked questions

What happens when a budget is exceeded?

Zelyx returns HTTP 402 before the provider is contacted, so no tokens are consumed and no cost is incurred. Alerts fire at 80%.

Which providers can be connected?

Supported providers are OpenAI, Anthropic, Google, Groq, Together, Mistral, DeepSeek, xAI, OpenRouter, Fireworks, Perplexity, or any custom slug. Supported providers also include Azure AI Foundry and Cohere.

Does the app see the provider key?

No. The provider key is stored encrypted in the vault and decrypted by the proxy. The app never sees the provider key. The Zelyx key is shown once and cannot be retrieved again.

How much of the SDK has to change?

Two lines: the key and the base URL. Model names, streaming, tool use, and message format stay the same. There is no SDK swap and no config file.

Is there a free start?

Zelyx is in private beta with design partners. Individual is free for 3 months for design partners, and the about page says to start on Free.

What does the gate check?

The gate checks budgets, model policies, and payment patterns, then forwards the request if it is approved. Caps on one provider do not protect fallbacks, which is why the budget sits in the call path.