Skip to content
NewReasoning, vision & image models on one key

One API key.
Every model you need.

Chat, reasoning, vision, embeddings and image generation behind a single OpenAI-compatible endpoint. Start free, scale on simple token plans, and watch every request in real time.

  • Free tier, no card
  • Works with OpenAI SDKs
  • Pay in USDT or PKR
output · deepseek-v4.1-flash
$

OpenAI SDK compatible — change base_url and your key. Nothing else.

Loading the model catalog…

Works with your tools

Bring any tool. Reach every model.

Point Claude Code, Cursor, Cline, Roo Code, Kilo Code, opencode or any OpenAI or Anthropic SDK at one base URL. Every request is routed to a healthy provider and billed to one key.

Cursor
Claude Code
Cline
Roo Code
Kilo Code
opencode
OpenAI SDK
FZ-craft

fzcraft.pikonox.com/v1

Live

POST /v1/chat/completions

fromCursor

modeldeepseek-v4.1-flash

reply200 OK · 180ms · 640 tokens

One key30+ modelsOpenAI + Anthropic APIAutomatic failover
Claude Fable 5$16.61
GPT-5.5$8.95
Gemini 3.1 Pro$4.47
DeepSeek V4.1 Flash$0.056200 · 180ms
Kimi K3$4.98
Qwen3.7 Max$1.76
Grok 4.6$2.81

Price per 1M input tokens

Platform

Everything between your code and the model

Routing, keys, limits, analytics and billing are handled for you — so shipping a new model is a one-line change, not a new integration.

One key, every model

A single OpenAI-compatible /v1 endpoint for chat, reasoning, vision, embeddings and images. Switching models is a one-string change.

-model="deepseek-v4.1-flash"
+model="claude-sonnet-5-5"
# same client, same code200 OK · 180ms
/chat/completions/embeddings/images/generations/images/edits/models

Live usage analytics

Tokens, requests, latency and cost per key and per model — updated as requests land, with full request logs.

Live

Requests

12,480

Tokens

8.2M

p95

210ms

Per-key spend caps & IP allowlists

Give every app its own key with scopes, a request-rate cap, token and spend limits, model and IP allowlists, and an expiry date.

fz_live_••••3f9a active
Spend$3.12 / $5.00
Requests this minute87 / 120
192.0.2.55blocked · 403
203.0.113.4allowed

Image Studio with its own wallet

Generate and edit up to 4K with style presets and any aspect ratio. Images are paid per image from a separate image wallet — failed images are never charged.

a nebula shaped like a phoenix · cinematic · 2KGenerate
Image wallet $2.00001K · 2K · 4K16:9 · 1:1 · 9:16

A Playground that thinks out loud

Test any model in the browser with visible reasoning, live web search with sources, and artifacts you can preview, copy or download.

Compare these two pricing pages and build a summary chart
Thought for 6sSearched · 4 sources Artifact · chart.html

Plan B is 18% cheaper per 1M tokens and adds vision. Here's a chart comparing both.

Pay the way you already pay

Crypto is verified automatically; local PKR transfers are reviewed by a person and credited to your balance.

Binance PayUSDT · BEP-20JazzCashEasypaisaBank transfer
Binance Pay · Plan activatedauto

Models

Pick a model. Keep your code.

Every model speaks the same API — OpenAI or Anthropic format. Pick a provider, copy the model id, and you're done.

Browse all models

Compare prices, filter by capability and estimate your monthly usage.

How it works

From sign-up to first token in three steps

  1. Create a free account

    Sign up with email or Google. The free plan includes a monthly token allowance and the free models — no payment details.

    you@company.com
    Create account Free plan · 1M tokens
  2. Create an API key

    Name it, choose scopes and optional limits. The key is shown once — store it as a secret.

    fz_live_8kQ2…f3a9 Copied
    inference120 req/minno expiry
  3. Point your SDK at us

    Set the base URL and key in any OpenAI or Anthropic SDK, framework or tool. Your existing code keeps working.

    base_url="https://fzcraft.pikonox.com/v1"

    200 OK · first token in 312ms

    Hello! How can I help today?

0:47from sign-up to your first token

Pricing

Simple plans, generous tokens

Start free. Upgrade when you need more tokens, higher rate limits or paid models. Yearly billing saves more.

Tokens, not surprises

Every plan is a fixed token allowance that refreshes every 30 days. Input and output tokens count; cached input counts at 10%.

Top up with a 24h pass

Ran out early? A pass adds tokens instantly and is used before your plan's allowance.

Images bill separately

Image generations draw from a separate image wallet, per image. Images that fail are never charged.

FAQ

Questions, answered

Can't find what you need? The documentation covers every endpoint, limit and error code.

Does it work with OpenAI and Anthropic SDKs?
Yes, both. /v1/chat/completions, /v1/embeddings, /v1/images/generations, /v1/images/edits and /v1/models follow the OpenAI API, and /v1/messages follows Anthropic's Messages API — including streaming, tool use and thinking. Every chat model works through either one, so the OpenAI SDK, the Anthropic SDK and Claude Code all work with the same key.
What does the free tier include?
A token allowance that refreshes every 30 days, the free chat and embedding models, streaming, tools and vision. You don't need to add any payment details to start.
What happens when I run out of tokens?
Requests return 429 with the code insufficient_quota until your allowance refreshes. You can upgrade your plan at any time, or buy a 24h pass — its tokens are available immediately and are used before your plan's allowance.
How are images billed?
Per image, from a separate image wallet, at a price set by the model and the quality tier (1K, 2K or 4K). The cost is held when you send the request and refunded for any image that doesn't come back, so failed images are never charged.
Which payment methods can I use?
Binance Pay and USDT on BNB Smart Chain (BEP-20) are verified automatically. JazzCash, Easypaisa and bank transfers are paid in PKR and reviewed manually before your balance is credited.
Can I limit what a key can do?
Yes. Every key can have its own scopes, requests-per-minute cap, token limit, spend cap, model allowlist, IP allowlist and expiry date. Keys can be paused or revoked instantly from the dashboard.

Your first request is a minute away

Create a free account, copy your key, and send a chat completion with the SDK you already use.