AI models & services

AI for your work.One account. One balance.

Chat, code, create media and extract data from documents. Use our browser workspaces, connect your tools or build with models and specialized services through one API.

Pay as you goClear costsNo subscription required.

Models · AI services · one key

See what you can doImages, documents, text and speech.
See what you can doExplore the possibilities

From invoice to usable data.

01 Your document
invoice_q3.pdf4 pages
02 Structured fields
1Vendor
Acme Corp
2Invoice
INV-3391
3Invoice total
$4,820.00
Document Analysis (Invoice)

Explore Primex

Use AI your way.Use AI your way.

Start in your browser, or connect the tools and applications you already use.

Your browser work and API usage draw from the same AI balance.

Routing

Two routes. Different priorities.

One API, with routing choices, usage records and final charges.

KNOWN INFRASTRUCTURE

Verified

Available today

A defined path for your requests.

Review the selected offering’s availability, limits and data-handling terms.

COST-FOCUSED ROUTING

Economy

Coming soon

Put cost first for everyday work.

A planned cost-focused option for experiments and everyday workloads. Check Catalog for availability.

Developer quickstart

Your tools. Connected to Primex.

Connect your agents, editors and workflows. Choose a tool for its setup guide.

Compatibility is endpoint- and offering-specific. The live Catalog and docs are authoritative.

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: "prx-...",
  baseURL: "https://gateway.primexos.com/v1"
});

const response = await client.responses.create({
  model: "YOUR_MODEL_ID",
  input: "Build something useful."
});

Keep your compatible SDK. Update your API key, base URL and model. Migration guide

  1. 01

    Get your API key

    Create an account and generate a prx-... key.

  2. 02

    Try a request

    Run it in Playground or connect your existing SDK.

  3. 03

    Add credits when ready

    Top up your AI balance and pay for measured usage.

Usage & spending

See what you use. Know what you pay.

See what each request used, including cache activity, how long it took and what you paid.

  • Request-level usage

    See the model or service, measured units, request time and final net cost, attributed to the key that sent the request.

  • Caching, accounted for

    See which input was reused from cache or written to it, alongside the recorded charge. Cache availability and pricing depend on the model.

  • Spending controls

    Requests that fail without usable output aren’t billed. Set monthly spending caps for your API keys and track costs in Usage. See the FAQ for partial-output billing and how caps work.

EXAMPLE USAGEreq_7f3a2c9e
Status
Succeeded
Model
gpt-5.6-sol
Routing
Auto → Verified
Tokens
8,192 in318 out
Input & cache
2,048 regular4,096 read2,048 write
Request time
2.31 s
Net cost
$0.033038

Live models & pricing

Start with what you need. Pay for what you use.

A few useful starting points from the live Catalog. One AI balance covers models and services, each priced in the unit it measures. No subscription required.

View all models & pricing

Open the Catalog for current availability and pricing.

Models and services, 1 of 2

Frequently asked questions

Getting started, choosing your tools and paying for your work.

Why Primex?

Primex brings models from different providers and services for images, video, speech, documents, search and more into one account and AI balance. Work in your browser or connect supported developer tools through OpenAI-compatible and Anthropic-compatible APIs. You don’t need separate provider accounts for Primex-managed services. Pay as you go, see what each request costs and set spending caps for your API keys. Verified routing gives your requests a defined infrastructure path, and our support team can help with setup, requests and billing.

Can I use Primex without coding?

Yes. Create a Primex account and open Chat for conversations, Studio for images and video, or Audio for voice and sound. You don’t need separate provider accounts for Primex-managed services.

For other tasks, the Catalog shows available models and services. Developers can try the Playground or create an API key. Telegram is optional.

Can I connect my app or coding tools?

Primex supports OpenAI-compatible and Anthropic-compatible APIs on supported models and endpoints. Use the same Primex key with tools such as Claude Code, Codex and Cursor, following the guide for your tool.

Features vary by model, endpoint and client. Check the setup guides and Catalog before connecting your workflow.

How does pricing work?

No subscription is required. Add AI credit and pay for measured usage. The Catalog shows your final rates, including input, output, cache reads and cache writes where supported. Start with our task examples and cost calculator.

Cache support matters: a lower headline token price can still cost more if reused input is billed at the full rate. Economy is coming soon; we plan discounts on most models and the same clear pricing.

Creating an account is free; AI usage is paid. Any welcome credit has eligibility and usage conditions shown in Billing. Your final charges appear in Usage.

How can I pay?

You can add AI credit with USDT or USDC on the networks shown in Billing.

Choose an amount in USD in Billing. Review the available methods, purchase limits, payment amount and credit you’ll receive before you pay. Options and payment currency can vary by method and order.

Questions about fees or which method to use? Contact support before paying. If you’ve paid but your credit hasn’t arrived, include your payment reference so we can check it.

Read the crypto and local payment guide

What if a request fails or I reach my spending cap?

A request that fails without delivering usable output isn’t billed. If it delivers usable output before a later error, that usage may be charged. Usage shows the outcome and charge. If something looks wrong, send support the request ID so we can investigate.

You can set a monthly spending cap for each API key. Requests already running can finish above the cap, so it isn’t a strict cutoff for work in progress.

How does routing work?

Choose from the routing options published for your selected model or service. Verified and Economy availability is listed in the Catalog. Current requests use Auto. Usage shows the served tier and final charge.

What happens to my data?

The provider serving your request receives the inputs it needs to process it. Ordinary direct-API prompts and successful responses aren’t stored in the Usage ledger. We retain billing and operational records, and error records may be kept for troubleshooting.

Chat saves conversations. Studio, Audio and uploaded files have their own storage rules, and upstream providers’ policies also apply. See our Privacy Policy for details.

Start with Primex

Your next idea starts here.

Build with the API, or use Chat, Studio and Audio in Primex AI. Start with one Primex account.

No credit card required to sign up. No subscription required.