All models
Qwen

Qwen3.8 MAX API: pricing, base URL and quick setup

Qwen3.8 MAX is available on A2Agent as qwen3.8-max at $2.00 per 1M input tokens, with a 1M context window, through one OpenAI-compatible endpoint. The strongest Qwen tier for reasoning, at a 1M window.

Specifications and pricing

Model idqwen3.8-max
PlatformQwen
Context window1M
Tagged forchat, reasoning
Discount…
Input / 1M tokens$2.00
Output / 1M tokens$6.00

Prices synced with platform billing (USD per 1M tokens).

Quick setup

curl https://a2agent.me/v1/chat/completions \
  -H "content-type: application/json" \
  -H "authorization: Bearer sk-your-a2agent-key" \
  -d '{
    "model": "qwen3.8-max",
    "messages": [{"role": "user", "content": "Reply with OK"}]
  }'

Using it from a coding client

Best kept as the escalation model rather than the default. Point your client at a Flash release for routine turns and switch to MAX when a task stalls.

How it compares within its family

Qwen3.8 MAX is the most expensive model in the Qwen family here, at $2.00 per million input tokens and $6.00 output, and one of the two carrying the reasoning tag. The other, Qwen3.7 MAX, lists at $1.70 and $5.10, about fifteen percent less, with the same 1M window and the same tag. The premium buys the later release, not more context. Below the MAX pair the line drops sharply: Plus at $0.50 to $1.20 on input, Flash from $0.15. That is a more than tenfold spread inside one platform, which is reason enough to reserve MAX for the calls that actually need it. A retrieval step or a summarisation pass does not.

Other models on this platform

Questions

What is the context window?

1M tokens, the same as every Qwen model here.

Is it worth the premium over Qwen3.7 MAX?

They share a window and a tag. Qwen3.8 MAX is the later release at about fifteen percent more on both sides.

Why is it so much more than the Flash tier?

Different tuning. The MAX models carry the reasoning tag; Flash carries chat and vision at roughly a tenth of the input rate.

Can I mix models on one key?

Yes. One key covers every model in the catalogue, and switching is a change to the model string.

Trial credit on sign-up, billed per token, no minimum spend.

Get an API keyCompare all prices