Alibaba Cloud
Qwen3 8 2 4t A95b on AtmoRouter — from $0.005 per million input tokens, from $0.015 per million output tokens on a 1M context, through one OpenAI-compatible key.
Input · per Mtok
From $0.005
$2.00 official
Output · per Mtok
From $0.015
$6.00 official
Off official
99.8%
Context
1M
up to 131K output
Final prices in USD per million tokens. Each tier is compared with official input and output prices.
| Input | Output | Offers |
|---|---|---|
| $0.00599.7% off official | $0.01599.7% off official | 20137 |
| $0.0597.5% off official | $0.1597.5% off official | 8875 |
| $0.1493% off official | $0.4293% off official | 214 |
| $0.3582.5% off official | $1.0582.5% off official | 73 |
| $1.1940.5% off official | $3.5740.5% off official | 1643 |
| $1.19540.2% off official | $3.58540.2% off official | 782 |
| $1.2537.5% off official | $3.7537.5% off official | 1406 |
| $2.33516.7% above official | $7.00516.7% above official | 107 |
| $2.525% above official | $7.525% above official | 1471 |
Offer counts are a snapshot and do not guarantee available capacity.
Updated
Prices vary by available offers. Requests can use any listed tier by default. Set maximum input and output prices in Budgets to limit routing.
Model id
ali/qwen3.8-2.4t-a95b
Use this exact id in API calls.
Modality
text
Prompt caching supported — cached input $0.0005/Mtok.
Protocols
OpenAI Chat Completions
Anthropic Messages endpoint also available for every model.
curl
curl https://atmorouter.dev/v1/chat/completions \
-H "Authorization: Bearer ar-xxxxxxxxxxxx" \
-H "Content-Type: application/json" \
-d '{
"model": "ali/qwen3.8-2.4t-a95b",
"messages": [{"role": "user", "content": "Say hello in one word."}]
}'Python (OpenAI SDK)
from openai import OpenAI
client = OpenAI(
base_url="https://atmorouter.dev/v1",
api_key="ar-xxxxxxxxxxxx",
)
resp = client.chat.completions.create(
model="ali/qwen3.8-2.4t-a95b",
messages=[{"role": "user", "content": "Say hello in one word."}],
)
print(resp.choices[0].message.content)Works with anything OpenAI-compatible — Cursor, Cline, LangChain and more. See the cookbook configs at github.com/Micolaabdi/atmorouter-cookbook and the docs for the Anthropic Messages endpoint.
How much does Qwen3 8 2 4t A95b cost per million tokens?
AtmoRouter prices start from $0.005 per million input tokens and from $0.015 per million output tokens. The provider's official list rate is $2.00 / $6.00 — AtmoRouter is 99.8% below it.
Does Qwen3 8 2 4t A95b support streaming?
Yes — server-sent events on the OpenAI-compatible endpoint, and the Anthropic Messages endpoint streams natively.
Is there a subscription or minimum spend?
No subscription or minimum commitment. Billing uses the selected offer’s rates, and a minimum request charge applies. See Budgets for the current minimum.