Catalog

Every model, one endpoint

Every dinar price is calculated in front of you: Alibaba's public dollar price, multiplied by 250 DA, plus our 20%. Nothing hidden and nothing inflated.

19 models shown · Models marked "public price verified" show the rate published by Alibaba Cloud. For the others, Alibaba does not publish a fixed rate (it depends on region and volume): the exact price appears in your console, with the same +20% rule.

qwen3.8-max

Qwen · Alibaba

Top model Public price verified

Flagship reasoning and agentic workhorse

Flagship Context 131 072

Input
600 DA
$2 Alibaba × 250 × 1,2
Output
1 800 DA
$6 Alibaba × 250 × 1,2

per 1M tokens

qwen3.7-plus

Qwen · Alibaba

Best value Public price verified

Best quality-per-dinar for production apps

Flagship Context 131 072

Input
82.8 DA
$0.276 Alibaba × 250 × 1,2
Output
330.3 DA
$1.101 Alibaba × 250 × 1,2

per 1M tokens

qwen3.8-flash

Qwen · Alibaba

Lowest cost Public price verified

High-volume chat, classification and extraction

Fast & cheap Context 131 072

Input
45 DA
$0.15 Alibaba × 250 × 1,2
Output
141 DA
$0.47 Alibaba × 250 × 1,2

per 1M tokens

qwen3.5-flash

Qwen · Alibaba

Public price verified

The cheapest model in the catalog

Fast & cheap Context 131 072

Input
30 DA
$0.1 Alibaba × 250 × 1,2
Output
120 DA
$0.4 Alibaba × 250 × 1,2

per 1M tokens

qwen3.5-plus

Qwen · Alibaba

Public price verified

A good step up when Flash is not enough

Fast & cheap Context 131 072

Input
34.5 DA
$0.115 Alibaba × 250 × 1,2
Output
206.4 DA
$0.688 Alibaba × 250 × 1,2

per 1M tokens

qwen-image-3.0-pro

Qwen · Alibaba

Public price verified

Commercial image generation and editing

Image

per generated image
22.5 DA
$0.075 Alibaba × 250 × 1,2

per generated image

deepseek-v4-pro-0813

DeepSeek

Deep reasoning

Long-horizon reasoning and code generation

Reasoning Context 262 144

per 1M tokens
Rate shown in your console

per 1M tokens

deepseek-v4.1-flash

DeepSeek

Fast reasoning at a fraction of the cost

Reasoning Context 131 072

per 1M tokens
Rate shown in your console

per 1M tokens

glm-5.2

Zhipu AI

Strong bilingual reasoning, Chinese + English

Flagship Context 131 072

per 1M tokens
Rate shown in your console

per 1M tokens

kimi-k3

Moonshot

Long context

256K context for document-heavy workloads

Flagship Context 262 144

per 1M tokens
Rate shown in your console

per 1M tokens

MiniMax-M2.5

MiniMax

Low-latency multilingual chat

Fast & cheap Context 131 072

per 1M tokens
Rate shown in your console

per 1M tokens

qwen3.8-omni-flash

Qwen · Alibaba

Multimodal

One model for image, audio and video understanding

Vision & Omni Context 131 072

per 1M tokens
Rate shown in your console

per 1M tokens

wan2.7-image-pro

Wan · Alibaba

Poster and product visuals at volume

Image

per generated image
Rate shown in your console

per generated image

wan3.0-video

Wan · Alibaba

Text-to-video and image-to-video generation

Video

per second of video
Rate shown in your console

per second of video

happyhorse-1.1-i2v

HappyHorse

Animate a still image into short video

Video

per second of video
Rate shown in your console

per second of video

qwen-audio-3.1-asr-flash-filetrans

Qwen · Alibaba

Transcribe recorded audio, 34 languages including Arabic

Audio & speech

per minute of audio
Rate shown in your console

per minute of audio

qwen-audio-3.0-tts-plus

Qwen · Alibaba

Natural Arabic, French and English voices

Audio & speech

per 1,000 characters
Rate shown in your console

per 1,000 characters

text-embedding-v4

Qwen · Alibaba

Vector search and retrieval for RAG pipelines

Embeddings & RAG Context 8 192

per 1M tokens
Rate shown in your console

per 1M tokens

qwen3-rerank

Qwen · Alibaba

Re-order search results for better answers

Embeddings & RAG Context 4 000

per document scored
Rate shown in your console

per document scored

Included with every key

What you get on top of the tokens

OpenAI & Anthropic compatibility

Two endpoints, same key. Keep your existing SDK, tools and prompts.

Per-key rate limits and whitelists

Set the allowed rate and permitted IP addresses, key by key.

Streaming, tools and structured output

Function calling, JSON mode, context caching and batch inference.

Top-up and tracking in dinar

Spending history, balance and invoices in your console, not someone else's.

Need a model that is not in the list?

Alibaba Cloud Model Studio publishes more than 150 models. Tell us which one and we will enable it in your console, on the same +20% rule.