Qwen API · Alibaba · OpenAI-compatible
Qwen3.8 Max API from $1.90 per 1M input tokens.
Alibaba’s Qwen3.8 models on one OpenAI-compatible key: 5% below Alibaba Cloud Model Studio’s International list price from your first request, and up to 25% below as your top-ups grow. No Alibaba Cloud account needed.
- $1 free when you sign up
- +10% on your first top-up, up to $100
- Failed requests are free
Qwen3.8 Max · per 1M input tokens
Official $2.00
$1.50Tier 3
Tier 1 for new accounts: $1.90. The more you top up, the lower it goes.
- Qwen3.8 Flash
$0.15$0.113−25%
What a top-up buys
$20 is about 3,000 Qwen3.8 Max chat replies.
Pick a model and an amount. Each number is how many requests of that size your balance covers at Tier 1, before the first top-up’s 10% extra credit.
What you pay for
A discount that grows, a long context, one key.
The more you top up, the less you pay
5% below Alibaba’s International list price at Tier 1, 15–20% at $1,000 in top-ups and 25% at $10,000. Every later request bills at the lower rate.
Long context at one price
Both models read up to 984K tokens, with text, image and video input, reasoning and tool calling. The price per token is the same at any prompt length.
Every major model family
No Alibaba Cloud account, region or console to set up. The same key reaches DeepSeek, GLM, Kimi, Claude, GPT and Gemini, from one prepaid balance in USD.
How pricing tiers work
Every account starts at Tier 1. Top up more to move up.
Your account starts on Tier 1, with the discount below from the first request. At $1,000 in cumulative top-ups it moves to Tier 2, at $10,000 to Tier 3, on its own, and every later request bills at the lower rate.
Qwen3.8 Max −5%, Qwen3.8 Flash −5% against Alibaba Cloud’s list price. Purchased credits last 12 months.
Qwen3.8 Max −15%, Qwen3.8 Flash −20%.
The lowest rates: Qwen3.8 Max −25%, Qwen3.8 Flash −25%.
- 20% or more
- 11–19%
- 10% or less
Swipe the table for Tier 2 and 3 →
| Per 1M tokens | Alibaba list price | Tier 1 | Tier 2 | Tier 3 |
|---|---|---|---|---|
| Qwen3.8 Max · input | $2.00 | $1.90−5% | $1.70−15% | $1.50−25% |
| Qwen3.8 Max · output | $6.00 | $5.70−5% | $5.10−15% | $4.50−25% |
| Qwen3.8 Flash · input | $0.15 | $0.143−5% | $0.12−20% | $0.113−25% |
| Qwen3.8 Flash · output | $0.47 | $0.447−5% | $0.376−20% | $0.353−25% |
Discounts apply to input and output. Cached input bills at Alibaba Cloud’s list price at every tier (Qwen3.8 Max $0.25, Qwen3.8 Flash $0.016 per 1M). Official price: the model maker's list price. Tier 1–3: OceanToken's standing discounts by cumulative top-up, not limited-time offers. Large volume? Talk to sales.
Which Qwen
Two Qwen models on the same key.
Both take up to 984K tokens of context, with reasoning, tool calling and text, image and video input. Pick by model ID.
Qwen3.8 Max
Alibaba’s largest Qwen3.8 model, for complex reasoning and coding. We serve the September 2 snapshot, up to 131K output tokens.
qwen/qwen3.8-max-0902
Qwen3.8 Flash
For high-volume work: about 13x cheaper than Max per input token, with the same context length, up to 131K output tokens.
qwen/qwen3.8-flash
Thinking mode
Alibaba prices Qwen3.8 Max the same with thinking on or off, and so do we.
thinking on · off
Setup
From sign-up to first request in a few minutes.
Questions
Before you top up
What do I get when I sign up?
An API key and $1 of free credit that lasts 14 days, no card needed: about 150 Qwen3.8 Max or 2,000 Qwen3.8 Flash chat replies of 2,000 input and 500 output tokens. Your first card top-up gets 10% extra credit, up to $100. Have a promo code? Redeem it in the console for more credit.
What does $20 get me?
At Tier 1, about 3,000 Qwen3.8 Max chat replies of 2,000 input and 500 output tokens, about 1,800 agent steps, or about 99 long documents of 100,000 input and 2,000 output tokens. On Qwen3.8 Flash, about 39,400 of the same chat replies. As a first top-up it buys 10% more.
How do your prices compare with Alibaba Cloud’s own API?
Against Alibaba Cloud Model Studio’s International list price, Qwen3.8 Max is −5% at Tier 1 ($1.90 input and $5.70 output per 1M, against $2.00 and $6.00) and Qwen3.8 Flash −5%. At Tier 3, Qwen3.8 Max is −25% and Qwen3.8 Flash −25%. Cached input bills at Alibaba’s list price.
Is Alibaba Cloud’s own price lower?
Sometimes. We compare with Alibaba Cloud Model Studio’s International list price. Alibaba also sells Qwen through its Global deployment at $1.65 input and $4.951 output per 1M for Max: below our Tier 1, below our Tier 2, above our Tier 3. Qwen3.8 Flash there is $0.113 and $0.382: below our Tier 1, below our Tier 2, the same as our Tier 3. Alibaba also offers batch requests at half price, which we do not.
With us you get one key and one USD balance for Qwen and every other model, without an Alibaba Cloud account or region to set up.
When do I move to Tier 2?
When your cumulative top-ups reach $1,000; Tier 3 starts at $10,000. Gifts and credit grants count toward the total; the first top-up’s extra credit does not. Your account moves up on its own and every later request bills at the lower rate; the console shows the rate that applies to you. Expect large volume? Talk to sales.
Is there a subscription?
No. You start on the free sign-up credit, then prepay a balance in USD by card, any amount, and each request is deducted from it. Purchased credits last 12 months.
Do I need to change my code?
Only the base URL and the key. The API is OpenAI-compatible, and the Anthropic Messages format works too.
What happens when my balance reaches $0?
New requests are refused until you add credits. Your key and your code stay the same.
More on OceanToken
Every Qwen model, side by side.
- Qwen3.8 Max Model page: prices, context, capabilities
- Qwen3.8 Flash Model page
- All Qwen models Every Qwen model on OceanToken with its price
- All models and prices Every model on the same key
Ready