NURA NURAChat · Private Inference
API Login
Pricing

Simple, competitive pricing

Pay-as-you-go AI inference and GPU compute — up to 40% cheaper when you pay and stake with AIX.

Start chatting Get AIX Rent GPUs
Benchmarked against Venice.ai and OpenRouter · Settled in AIX on BSC
−20%
Pay with AIX
−40%
Up to, with staking
~70%
Revenue to GPU providers
Inference

Token pricing

USD per 1M tokens. ≈10–15% below OpenRouter list; pay with AIX for −20%.

Model Input $/1M Output $/1M Output w/ AIX −20%
DeepSeek V4 Flash 0.200.16 0.850.68 0.68
DeepSeek V4 Pro 0.450.36 1.751.40 1.40
Qwen3.6 Flash (Alibaba) 0.180.14 0.700.56 0.56
Qwen3.6 Plus (Alibaba) 0.400.32 1.601.28 1.28
Grok 4.x (xAI) 2.502.00 12.009.60 9.60
Claude Sonnet 2.802.24 14.0011.20 11.20
GPT-5.x (OpenAI) 2.201.76 9.007.20 7.20
Open models on NURA workers (Llama / Qwen OSS) 0.150.12 0.450.36 0.36

Indicative list rates; actual cost varies with load and context length. Benchmarked against OpenRouter and Venice.ai, July 2026.

Images

Image generation

Priced per image, up to 2048². Image-to-image supported.

Qwen-Image / Wan
$0.03 / image
Up to 2048² · txt2img + img2img
Grok Imagine
$0.02 / image
Fast, stylized generations
Venice / Flux
$0.04 / image
High-detail photoreal & art
GPT-Image
$0.08 / image
Premium quality & text rendering
GPU compute

Rent GPUs by the hour

Dedicated GPUs from NURA providers. Prices shown "from", per hour.

GPU Spec Price
RTX 409024 GB · best valuefrom $0.30 /hr
RTX 509032 GB · next-genfrom $0.45 /hr
L40S48 GB · inferencefrom $0.90 /hr
A100 80G80 GB · trainingfrom $1.20 /hr
H10080 GB · top tierfrom $2.20 /hr

Browse the marketplace →

Plans

Access tiers

Start free. Upgrade for API access, higher limits, and monthly credit.

Free
$0always free
  • Chat on open models
  • 200 messages / day
  • Community rate limits
  • No API key
Start chatting
Most popular
Pro
$18/mo
  • All models, pay-as-you-go at list
  • API access
  • Higher rate limits
  • $5 monthly inference credit
  • Image generation
Choose Pro
Scale
$99/mo
  • Volume discounts
  • Priority routing
  • Team seats
  • $30 monthly credit
  • SLA support
Choose Scale

Every plan can pay with AIX for an extra −20%, stackable with staking tiers.

AIX token

How AIX powers NURA

AIX is NURA's utility token on BSC (BEP-20). Use it to pay, stake, and earn across the network.

01
Get AIX
Buy on BSC; hold in your wallet or NURA balance.
02
Pay & stake
Pay with AIX for −20%; stake for up to +20% more.
03
Metered deduction
Every token, image, and GPU-hour deducted at your rate.
04
Network earns
Providers, stakers, and referrers are paid in AIX.
Rewards recirculate

Pay with AIX → 20% off

Settle any invoice from your AIX balance and every list price drops 20% — automatically.

Example
DeepSeek V4 Pro · 1M output tokens $1.75 → $1.40 save $0.35
H100 · 1 hour $2.20 → $1.76 save $0.44

Stake AIX → tier discount (stacks, capped at 40%)

Lock AIX to unlock a routing tier. The staking discount stacks on top of the 20% pay-with-AIX discount — up to 40% off in total.

Bronze
stake 1,000 AIX
+5% off
Standard routing
Silver
stake 10,000 AIX
+10% off
Faster routing
Gold
stake 50,000 AIX
+15% off
Priority routing + higher rate limits
Platinum
stake 200,000 AIX
+20% off
Top priority + dedicated capacity + early access
20% pay + 20% Platinum stake = 40% max discount

Stakers also earn a share of network fees (variable APR).

Transparent metering

Usage is metered per token, per image, and per GPU-hour, then auto-deducted from your balance at your effective (discounted) rate. Every request returns a line item in both USD and AIX.

Sample line item
DeepSeek V4 Pro · 1,240 output tokens
effective rate$1.40 / 1M
charged$0.00174 ≈ 0.62 AIX

Earn AIX

Run a GPU
Providers run the NURA host agent and earn ~70% of the inference revenue their GPU serves — in AIX — plus uptime rewards.
Stake
Stakers earn network-fee APR proportional to their stake.
Refer
Referrers earn 10% of referred spend for 6 months.

How the economics work

NURA keeps ~30% platform margin on relayed inference (funding the ~70% provider payout), and ~20% on NURA-worker inference (providers keep ~80%). AIX discounts are funded by routing demand to cheaper DePIN capacity plus token utility; staking locks supply and rewards holders.

Not financial advice; token utility only.
FAQ

Frequently asked

How am I billed?
Pay-as-you-go by default: usage is metered and deducted from your prepaid balance (card or AIX). Plans add monthly credit and higher limits.
Can I pay in fiat?
Yes. Top up your balance with a card or USDT, or pay with AIX for −20%. Fiat and AIX balances are billed at the same USD rates.
What's the AIX discount stack?
20% for paying with AIX, plus 5–20% for your staking tier — capped at 40% total off list.
How do I earn AIX?
Run a GPU with the NURA host agent (~70% of served revenue), stake for network-fee APR, or refer users for 10% of their spend for 6 months.
Is there a free tier?
Yes — Free gives you chat on open models with 200 messages/day at community rate limits, no card required.
How does per-request metering work?
Each request meters exact tokens, images, or GPU-seconds and deducts your effective (discounted) rate. You get a line item in USD and AIX every time.