AI Model Watch
Models / Alibaba

Qwen-Max (Qwen2.5-Max)

GA

Legacy stable Max alias = qwen-max-2025-01-25 (a.k.a. Qwen2.5-Max). Official International price: $1.6 in / $6.4 out per 1M, no tiered pricing, Non-Thinking mode only; supports 50% batch-inference discount. Context window ~32,768 (33K) and max output ~8,192 per third-party corroboration (CloudPrice/DataStudios); not on accessible official spec page. Still listed/purchasable on pricing page (Jun 22, 2026); not in current deprecation table. Released 2025-01-27 (International). API-only/proprietary.

Qwen-Max (Qwen2.5-Max) by Alibaba costs $1.60 per 1M input tokens and $6.40 per 1M output tokens ($2.80/1M blended), with a 33K (32.768-token) context window. It is generally available (GA).

Last verified: 13 Aug 2026 · sourced from official provider documentation

Provider
Alibaba
Status
GA
Input price
$1.60 / 1M tokens
Output price
$6.40 / 1M tokens
Cached input
—
Blended price
$2.80 / 1M tokens
Context window
32.768 tokens (33K)
Max output
8192 tokens
Modality
text
Open weights
No — API only
Knowledge cutoff
—
Released
27 Jan 2025
API string
qwen-max

Source: Alibaba official documentation ↗

See also

Compare Qwen-Max (Qwen2.5-Max) head-to-head

Qwen-Max (Qwen2.5-Max) vs Claude Opus 4.8→

$2.80 vs $10 blended /M · cross-provider

Qwen-Max (Qwen2.5-Max) vs GPT-5.6 Sol→

$2.80 vs $11.25 blended /M · cross-provider

Qwen-Max (Qwen2.5-Max) vs Gemini 3.6 Flash→

$2.80 vs $3 blended /M · cross-provider

Qwen-Max (Qwen2.5-Max) vs Grok 4.5→

$2.80 vs $3 blended /M · cross-provider

Qwen-Max (Qwen2.5-Max) vs Mistral Large 3→

$2.80 vs $0.75 blended /M · cross-provider

Qwen-Max (Qwen2.5-Max) vs Amazon Nova Pro→

$2.80 vs $1.40 blended /M · cross-provider

FAQ

Qwen-Max (Qwen2.5-Max) — questions & answers

How much does Qwen-Max (Qwen2.5-Max) cost?

Qwen-Max (Qwen2.5-Max) is priced at $1.60 per 1M input tokens and $6.40 per 1M output tokens ($2.80/1M blended at a 3:1 input-to-output ratio).

What is the context window of Qwen-Max (Qwen2.5-Max)?

Qwen-Max (Qwen2.5-Max) has a 32.768-token context window (33K), with up to 8192 output tokens per request.

Is Qwen-Max (Qwen2.5-Max) deprecated?

No — Qwen-Max (Qwen2.5-Max) is generally available (GA) and not currently scheduled for deprecation or retirement in our tracker.

Does Qwen-Max (Qwen2.5-Max) support image or vision input?

Our catalog lists Qwen-Max (Qwen2.5-Max)'s modalities as text; image/vision input is not among them.

Can I self-host Qwen-Max (Qwen2.5-Max)?

No — Qwen-Max (Qwen2.5-Max) is proprietary. Its weights are not publicly released, so it is available only through Alibaba's API (or hosted partners), not for self-hosting.

What is the API model string for Qwen-Max (Qwen2.5-Max)?

The API model identifier for Qwen-Max (Qwen2.5-Max) is "qwen-max" when calling Alibaba's API.