Qwen API pricing tiers: Max, Plus, Flash explained

Qwen API pricing tiers explained

Updated 2026-10-29. Official Alibaba Cloud prices, LMArena ranks.

Quick answer

Qwen's API runs on three price tiers: Max ($2 in and $6 out for Qwen3.8 Max), Plus ($0.40 / $1.60 for Qwen3.7 Plus) and Flash (from $0.03 / $0.13 for Qwen3.7 Flash), plus open-weight sizes between $0.25 and $2 input. The newer Qwen3.8 Max is cheaper than Qwen3.7 Max ($2.50 / $7.50) and ranks #10 on Code Arena against #30. Prices are Alibaba's international (Singapore) rates, and several tiers charge more once a prompt passes 32K or 256K tokens.

Model	Input	Output	Context	Long-prompt tiers
Max
Qwen3.8 Max	$2	$6	1M	single rate
Qwen3.7 Max	$2.50	$7.50	1M	single rate
Plus
Qwen3.7 Plus	$0.40	$1.60	1M	single rate
Qwen3.6 Plus	$0.50	$3	1M	above 256K: $2 in, $6 out
Qwen3.5 Plus	$0.40	$2.40	1M	above 256K: $0.50 in, $3 out
Flash
Qwen3.7 Flash	$0.03	$0.13	1M	32K to 256K: $0.10 in, $0.40 out; 256K to 1M: $0.20 in, $0.80 out
Qwen3.8 Flash	$0.15	$0.47	1M	single rate
Qwen3.5 Flash	$0.10	$0.40	1M	single rate
Qwen3.6 Flash	$0.25	$1.50	1M	above 256K: $1 in, $4 out
Coder
Qwen3 Coder Flash	$0.30	$1.50	1M	tiered, see note
Open-weight sizes
Qwen3.8 2.4T-A95B	$2	$6	1M	single rate
Qwen3.8 27B	$0.50	$3	1M	single rate
Qwen3.6 27B	$0.60	$3.60	256K	single rate
Qwen3.6 35B-A3B	$0.38	$2.25	256K	single rate
Qwen3.5 27B	$0.30	$2.40	256K	single rate
Qwen3.5 35B-A3B	$0.25	$2	256K	single rate
Qwen3.5 122B-A10B	$0.40	$3.20	256K	single rate
Qwen3.5 397B-A17B	$0.60	$3.60	256K	single rate

USD per 1M tokens, Alibaba Cloud Model Studio international (Singapore) list price, checked 2026-10-29. Long-prompt tiers are parsed from the notes on the official page; "see note" means the page describes tiers we show on the model page instead.

How Qwen's tiers work

Max is the flagship. Qwen3.8 Max is $2 in and $6 out with a 1M window, #10 on Code Arena.
Plus is the balanced tier. Qwen3.7 Plus lists at $0.40 in and $1.60 out, and Alibaba notes a limited-time 20% off on that list price.
Flash is the cheap tier. Qwen3.7 Flash is $0.03 in and $0.13 out, number 1 of 77 on our full price list; Qwen3.8 Flash is $0.15 / $0.47.
Open-weight sizes (27B up to 397B-A17B, plus a 2.4T-A95B model) have their own prices. The 3.5 and 3.6 sizes run $0.25 to $0.60 input with a 256K window; Qwen3.8 27B has a 1M window at $0.50 in.

The long-prompt catch

Qwen is the provider where prompt length changes the price most often. Qwen3.7 Flash charges 32K to 256K: $0.10 in, $0.40 out; 256K to 1M: $0.20 in, $0.80 out. Qwen3.6 Plus is $0.50 in and $3 out up to 256K and $2 / $6 above it. Qwen3 Coder Flash starts at $0.30 in for prompts up to 32K and climbs with length, to $1.60 in and $9.60 out above 256K. If your prompts are long, price the long tier, not the headline.

Cached input

Alibaba lists the cache-hit price only in the Model Studio console for Qwen3.8 Max, Qwen3.8 Flash and several others, so we show none for them. Qwen3.7 Plus lists $0.04 for explicit cache hits and $0.08 for implicit ones.

What a month costs

List price, no discounts, our math. Medium is 20M input and 5M output tokens a month; heavy is 200M and 50M.

Model	Medium month	Heavy month
Qwen3.8 Max	$70.00	$700
Qwen3.7 Plus	$16.00	$160
Qwen3.7 Flash	$1.25	$12.50
Qwen3.8 Flash	$5.35	$53.50

Qwen against the models people compare it with

Qwen3.8 Max vs GPT-6.1 Terra: $3 vs $2 blended, #10 vs #11 on Code Arena.
Qwen3.8 Max vs Claude Opus 5.5: 2.7x cheaper blended.
Qwen3.8 Max vs DeepSeek V4 Pro and Kimi K3.1 Pro vs Qwen3.8 Max.

Sources

FAQ

How much does the Qwen API cost?

Qwen3.8 Max costs $2 per 1M input tokens and $6 output. Qwen3.7 Flash is $0.03 and $0.13. All prices are Alibaba's international rates, as of 2026-10-29.

What are the Qwen model tiers?

Max for the strongest models, Plus for balanced use, Flash for cheap bulk work, a Coder model, and open-weight sizes from 27B parameters up.

Does Qwen charge more for long prompts?

On several models, yes. Qwen3.7 Flash charges $0.03 in up to 32K tokens, $0.10 from 32K to 256K and $0.20 from 256K to 1M.

Is Qwen3.8 Max cheaper than Qwen3.7 Max?

Yes. Qwen3.8 Max is $2 in and $6 out, against $2.50 and $7.50 for Qwen3.7 Max, and it ranks higher on Code Arena.

What is the cheapest Qwen model?

Qwen3.7 Flash at $0.03 input and $0.13 output per 1M tokens, the cheapest of all 77 models we track.

Are these Qwen prices for China or international?

International: Alibaba Cloud's Singapore region. Mainland China pricing is different.