New from Alibaba

Qwen3.8 Max online: Alibaba's largest model, pay per use

Qwen3.8 Max is Alibaba's biggest model yet: 2.4 trillion parameters in a mixture of experts, text, images and video in, a million tokens of context. It launched on August 3, 2026, and the September 2 snapshot sharpened coding on engineering-scale projects and long autonomous agent runs. On avots.ai you use Qwen3.8 Max and the budget Qwen3.8 Flash pay per use, no plan, and the same balance also covers GPT-6 Astra, Claude Fable 5.1, Gemini and 40+ image, video and music models.

Use Qwen3.8 now ✦ See pricing
Two tiers, one predecessor

Qwen3.8 Max, Qwen3.8 Flash and Qwen3.7 Max side by side

 Qwen3.8 MaxQwen3.8 FlashQwen3.7 Max
Best forbig coding projects, long agents, visionvolume work, documents, video analysisthe previous flagship
Price per 1M in / out$2 / $6$0.15 / $0.47$2 / $6
Context window1M tokens1M tokens1M tokens
Max output131k tokens131k tokens65k tokens
Inputtext, images, videotext, images, videotext
On avots.aipay per usepay per usepay per use

The number is the generation, the name is the tier: Max is the 2.4 trillion parameter flagship, Flash the fast multimodal tier at a thirteenth of the price. Both read video. Switch freely, one balance covers the whole Qwen family.

What is new

What Qwen3.8 brings

Alibaba's largest model

2.4 trillion parameters, previewed at the World AI Conference in Shanghai and released on August 3, 2026.

Video in, at both tiers

Max and Flash read images and video natively: describe a clip, pull facts from a recording, read charts and long documents.

The 0902 coding update

The September 2 snapshot targets engineering-scale software projects and longer autonomous development runs, with stronger vision.

1M context, 131k output

A whole codebase or a long video transcript in one conversation, and answers twice as long as Qwen3.7 allowed.

Three steps

How to use Qwen3.8

Create an account

Sign up on app.avots.ai with email, Google or Telegram. Takes a minute.

Top up from 0.99 EUR

1,000 tokens cost 0.99 EUR. No plan to choose, no renewal, the balance sits there until you use it.

Pick Max or Flash

Choose a Qwen3.8 tier in the model picker, or let Agent Avots route the task for you. Every reply shows its token cost.

Sign up and start
One balance

Where you can use Qwen3.8

Web app

app.avots.ai on desktop and mobile, no install needed.

Telegram

Qwen3.8 in @AvotsAIbot, same balance, from any phone.

MCP

Use it inside Claude Desktop, Cursor or Cline via the MCP server.

Try Qwen3.8 for under a euro

1,000 tokens cost 0.99 EUR and a Qwen3.8 Flash reply costs a few of them. The same balance unlocks Max, GPT-6 Astra, Claude Fable 5.1, Gemini, Veo video and Nano Banana images.

Sign up and chat ✦
Guides

Related pages

Qwen3.7

The previous generation: Max, Plus and Flash, still in the picker.

DeepSeek V4

Open-weight MoE, 1M context, top LiveCodeBench coding scores.

Kimi K3

Moonshot's frontier open model, number one at coding, 1M context.

GPT-6 Astra

OpenAI's new flagship and its Pro mode, pay per use.

avots.ai runs 40+ AI models for chat, image, video and audio on one balance. See the full list on the Tools page or open the app and start creating.

Browse all model pages Open the app
FAQ

Qwen3.8 questions

What is Qwen3.8 Max?
Qwen3.8 Max is Alibaba's largest and most capable model: a 2.4 trillion parameter mixture of experts that reads text, images and video and writes text, with a 1 million token context window and up to 131k tokens of output. It launched on August 3, 2026, and the 0902 snapshot of September 2 improved coding on engineering-scale projects, long autonomous agent runs and native vision.
What is Qwen3.8 Flash?
The fast tier of the same generation: a multimodal reasoning model for coding assistance, agent workflows, document and codebase analysis, chart reading and long-video analysis, at 0.15 USD per million input tokens and 0.47 per million output.
How much does Qwen3.8 Max cost on avots.ai?
Pay per use, no subscription. Qwen3.8 Max is priced at 2 USD per million input tokens and 6 USD per million output, a fifth of GPT-6 Astra; Qwen3.8 Flash at 0.15 and 0.47. On avots.ai you top up from 0.99 EUR and every reply shows its token cost before you run it.
Qwen3.8 or Qwen3.7?
3.8 Max is the bigger model with video input and the September coding update; 3.7 Max, Plus and Flash stay in the picker at their prices. For new work start with 3.8 Max or 3.8 Flash and keep 3.7 for prompts you have already tuned.
Can Qwen3.8 read a video?
Yes. Both Max and Flash take video as input and can describe, summarise or answer questions about a clip, next to images and long documents. Attach the file in the avots.ai chat and ask.
Try it on avots ✦
Start creating