OPENAI-COMPATIBLE API Access by quotation
AI Nexus logomark

One Hub for
All AI Models

GLM · DeepSeek · Seedream · Seedance

Access top-tier text, image and video models through a single OpenAI-compatible API. No re-platforming, no fragmented vendors — model capacity you consume like a utility.

Request access Browse models
10
Production models · added over time
3
Text / image / video
99.9%
Availability design target
24/7
Priority incident escalation

01 / CAPABILITIES

Core capabilities

Everything from onboarding to metering, closed inside one platform.

Multi-model aggregation

Call GLM, DeepSeek, Seedream and Seedance from one account. No separate vendor onboarding, no fragmented billing.

Drop-in unified API

OpenAI-compatible interfaces. Keep your SDK and your code — swap the base URL and the API key and you are migrated.

Usage-based billing

Text per token, images per output, video per token converted from generated duration. Pay for what you use, no minimum spend.

Volume programmes

Committed-volume and reseller programmes come with top-up credits; the larger your forecast, the better the rate we can quote.

Stable routing

Peak / off-peak scheduling plus context caching keep long sessions fast and costs predictable.

Data protection

Encrypted transport, keys stored independently with rotation on demand, and retention limits defined in our Privacy Policy.

02 / MODEL MATRIX

Model matrix

Three capability types, 10 production models, one endpoint for all of them.

TEXT

Text & chat

General chat, code, long-context understanding and reasoning

  • GLM-5.2Flagship reasoning · cost-efficient
  • GLM-5.3-FlashLightweight · everyday chat
  • DeepSeek-V4-proDeep reasoning · complex tasks
  • DeepSeek-V4-flashPeak / off-peak tiers
  • DeepSeek-V4.1-flashNext-gen flash · context caching

Peak / off-peak routing and context caching lower the cost of long sessions.

IMAGE

Image generation

Creative design, marketing assets and visual production

  • Seedream 4.0Mature and stable · batch output
  • Seedream 4.5Higher fidelity · refined detail
  • Seedream 5.0 seriesIncludes Lite / Pro

Seedream 5.0 is the current flagship: layer separation and high-resolution output, from draft to final asset.

VIDEO

Video generation

Short-form video, advertising and motion content

  • Seedance 2.0 miniLight and fast · cost first
  • Seedance 2.0 fastFastest turnaround · quality first

Text-to-video and image-to-video at 720p / 480p, billed per token and converted from generated duration.

Models are delivered through authorised upstream channels; availability and unit prices may vary by region and are confirmed in your quotation.

03 / BILLING

How billing works

One metering model across every provider, quoted in USD. We publish the mechanics here and the numbers in your quotation.

Text & chat models

Tokenper 1K tokens

Input and output metered separately; cache hits are billed at a lower rate.

Image models

Imageper output

Priced per generated image; high-resolution output carries no surcharge.

Video models

Token≈ per second

Billed per token and converted into a per-second equivalent before you commit volume.

/ TOKENText metered per token, input and output
/ IMAGEImages billed per generated output
/ SECONDVideo billed per token, converted from duration

Rates, discounts and top-up credits are stated in your individual quotation and on the settlement page. GLM-5.3-Flash is billed at upstream list price and is excluded from credit and volume tiers.

04 / PROGRAMMES

Volume & credit programmes

Three ways to buy capacity. Top-up credits sit in the same balance as paid credit and are spent with equal priority.

Pay as you go

No commitment

  • Top up, then consume; balance is transparent in the console
  • Built for evaluation and low-volume pilots
  • Every available model is open from day one
Request access
Most popular

Committed volume

Most requested

  • Discount negotiated against your forecast spend
  • Top-up credits included, validity stated in your quotation
  • Suits day-to-day production for small teams
  • Named technical support channel
Request a quote

Enterprise & reseller

Scale and resale

  • Deepest rates, tiered by committed spend
  • Bank transfer and consolidated invoicing
  • Sub-accounts and multi-key management
  • Routing and SLA terms available on request
Talk to sales

Credit amounts, validity periods and eligible models are defined in your quotation and in the Credit & Refund Policy.

Exception: GLM-5.3-Flash is sold at upstream list price and does not participate in credits or volume tiers. Image-model credits may be capped at a lower tier — your quotation states the applicable cap.

05 / HOW IT WORKS

Live in four steps

No rewrite of existing code. Point the endpoint at the hub and you are migrated.

  1. 01

    Request access

    Email your use case and expected volume; we open the account and send onboarding steps.

  2. 02

    Create API keys

    Issue keys in the console, keep them server-side, rotate them at any time.

  3. 03

    Swap the base URL

    Change base_url only. Model names and request shapes stay as they are.

  4. 04

    Meter and settle

    Usage is metered per token and per image in real time, with itemised invoices.

quickstart.py 
from openai import OpenAI

client = OpenAI(
    api_key="sk-your-api-key",
    base_url="https://api.useainexus.com/v1"  # change this line only
)

resp = client.chat.completions.create(
    model="glm-5.2",
    messages=[{"role": "user", "content": "Hi, tell me about yourself."}]
)
print(resp.choices[0].message.content)

06 / CONTACT

Get access

Tell us the workload and the volume. You get rates, a sandbox key and integration support.

Sales & partnerships

EMAIL

[email protected]

Commercial, channel and reseller enquiries. We aim to reply within one business day.

Technical support

EMAIL

[email protected]

Integration, latency and error-code questions. Priority incidents escalated 24/7.

Service hours

TIMEZONE

09:30 – 18:30 UTC+8, Mon–Fri

Outside these hours your email still enters the on-call queue.