Skip to main content
TokenCost logoTokenCost

Qwen4-Max

Last verified November 16, 2026 · Alibaba pricing ↗

$1.50/1M input · $4.50/1M output · 1.0M context · Alibaba

Qwen4-Max costs $1.50 per 1M input tokens and $4.50 per 1M output tokens from Alibaba, with a 1.0M-token context window. A typical chat reply (1,500 tokens in, 400 out) costs $0.0040, or $40.50 per 10,000 requests. On that workload it is cheaper than 70 of the 175 other buyable models tracked here.

Count Tokens

—
Tokens
—
Input Cost
—
Output Cost

Estimate Monthly Cost

Monthly Cost Estimator

Quick:
<$0.0001/mo

Pick a preset above or enter custom usage

What Qwen4-Max Costs Per Request

At $1.50 per 1M input and $4.50 per 1M output tokens, before caching or batch discounts.

WorkloadInput / output tokensPer requestPer 10,000 requests
Chat reply1,500 / 400$0.0040$40.50
RAG answer over retrieved docs8,000 / 600$0.015$147.00
Coding-agent step30,000 / 2,000$0.054$540.00

Qwen4-Max vs Similarly Priced Models

Nearest buyable models by price, with the RAG answer above (8,000 in / 600 out) priced on each.

ModelInput /1MOutput /1MContextRAG answer
Qwen4-Max Alibaba$1.50$4.501.0M$0.015
Qwen3.5-Omni Plus Alibaba$0.400$4.80262K$0.0061
GPT-6.1 Terra OpenAI$1.00$5.001.1M$0.011
Claude Haiku 4.5 Anthropic$1.00$5.00200K$0.011

Alternatives to Qwen4-Max

Pricing Details

AlibabaAPI id qwen4-max (OpenRouter qwen/qwen4-max, canonical qwen/qwen4-max-20261115, listed 2026-11-16 02:18 UTC). Closed weights; 2.4T backbone / 71B active plus 328B N-gram embedding parameters. Stored price is Model Studio Singapore (International): $1.50 input, $0.075 cache read (implicit or explicit, 5% of input), $1.875 5-minute cache write, $3.00 1-hour write, $4.50 output; 1M free tokens for 90 days. Global (HK/Frankfurt/Virginia/Tokyo) and Beijing: $1.413 / $0.071 / $1.767 / $2.827 / $4.24. One flat tier to 1M context, thinking and non-thinking billed the same, no batch price published. 25% under Qwen3.8-Max ($2/$6) on input and output, 70% under on cache read ($0.25). Max input 991,808 (983,616 thinking); max output 131,072 non-thinking, 262,144 thinking+answer per the model page, but QwenCloud and OpenRouter say 393,216. Thinking on by default, reasoning_effort low/high/xhigh, default high (preview page said xhigh). Text/image/video in. TPM 5M on QwenCloud. qwen4-max-preview was Token Plan only and ends 2026-11-25 24:00 UTC+8. The '$0.075 per million input tokens' figure in some coverage is the cache-read rate. Artificial Analysis v5 (2026-11-16, effort not stated): index 43, $3.27 per task, 127M output tokens vs 78.6M median ('very verbose'); Qwen3.8-Max 0902 is 34 at $10.02. Vendor-reported: Work360 52.69 (3.8: 47.74), cost per task $4.43 to $1.67; DeepSWE v1.1 75.7, Terminal-Bench 4.0 48.5 (Opus 5.5 59.6, Kimi K3.1 67.2), ProgramBench 31.5 (Opus 65.0), HLE 51.7.
Input / 1M tokens
$1.5
Output / 1M tokens
$4.5
Context Window
1.0M
Max Output
131K

Price History

Launched at current price on 2026-11-16. No price changes recorded since.

Frequently Asked Questions

Common questions about Qwen4-Max pricing and usage

How much does Qwen4-Max cost?
Qwen4-Max costs $1.5 per 1 million input tokens and $4.5 per 1 million output tokens. This pricing is set by Alibaba.
What tokenizer does Qwen4-Max use?
Qwen4-Max uses the cl100k_base tokenizer. This tokenizer has a vocabulary of ~100,000 tokens and is widely used across many models.
What is Qwen4-Max's context window?
Qwen4-Max supports a context window of 1.0M tokens (1,000,000 tokens). The maximum output length is 131K tokens.
How does Qwen4-Max compare to other models?
On an 8,000-token-in, 600-token-out RAG answer, Qwen4-Max costs $0.015 per request. The nearest-priced alternatives cost: Qwen3.5-Omni Plus (Alibaba) $0.0061, GPT-6.1 Terra (OpenAI) $0.011, Claude Haiku 4.5 (Anthropic) $0.011. The table above lists their full rates and context windows.
How do I count tokens for Qwen4-Max?
Use the token counter at the top of this page or visit the homepage calculator. Paste your text and it will show the exact token count and cost for Qwen4-Max using the cl100k_base tokenizer.

Read More About Qwen4-Max