Qwen 3.7 Max Pricing

Qwen 3.7 Max API pricing

Qwen 3.7 Max API pricing is $2.5 per million input tokens and $7.5 per million output tokens, verified November 4, 2026.

Input (per 1M tokens)

$2.5

Output (per 1M tokens)

$7.5

Cached input (per 1M tokens)

$0.25

Batch discount

50%

Context window

1,000,000 tokens

Max output

32,768 tokens

Capabilities

Qwen 3.7 Max supports image input, tool use, long context.

Qwen 3.7 Max API pricing has changed 3 times on record, most recently on October 10, 2026 (increase).

• October 10, 2026: Increase — input $1.25 → $2.5/M (+$1.25) — output $3.75 → $7.5/M (+$3.75)
• August 17, 2026: Decrease — input $2.5 → $1.25/M (−$1.25) — output $7.5 → $3.75/M (−$3.75)
• August 10, 2026: Increase — input $1.25 → $2.5/M (+$1.25) — output $3.75 → $7.5/M (+$3.75)

Frequently asked questions

Short answers to the most common questions about Qwen 3.7 Max pricing, limits, and capabilities.

Notes

Alibaba's flagship Chinese frontier model. List price $2.50/$7.50; currently on 50% promotional discount through Alibaba Cloud at $1.25/$3.75. Cached input drops to $0.25/M. Roughly half the price of Claude Opus on input. [2026-08-10] PROMO ENDED — Input rose $1.25 → $2.50 and Output rose $3.75 → $7.50 as the 50% launch promotion lapsed; now billing at list price per Alibaba Cloud Model Studio (corroborated by pricepertoken + eesel). Cache Read left at $0.25 — sources conflict on post-promo cache rate ($0.25 vs $0.50); verify next run. [2026-08-17] CORRECTION — THE 08-10 ENTRY WAS WRONG. The 50% promo did NOT end; it is still live and runs through 2026-08-31. Reverted Input $2.50 → $1.25 and Output $7.50 → $3.75. Confirmed against Alibaba's own promo page (help.aliyun.com/en/lingma/qwen3-7-max-limited-time-offer, alibabacloud.com/campaign/qwen-discount) plus docs.qoder.com, eesel.ai and pricepertoken — all report $1.25/$3.75 active with list at $2.50/$7.50. Root cause: the 08-10 run read list price as the effective rate. REAL EXPIRY TO WATCH: 2026-08-31 — on Sep 1 expect the genuine step up to $2.50/$7.50 unless extended. Cache Read still $0.25, unverified against post-promo cache rate. [2026-09-28] Verification attempt produced no clear result — recheck next run. IMPORTANT: the flagged promo expiry (2026-08-31) has now passed and no source this run could confirm whether pricing stepped up to list ($2.50/$7.50) or the promo was extended — prices LEFT UNCHANGED at $1.25/$3.75 pending direct confirmation. Prioritize checking Alibaba Cloud Model Studio pricing/promo pages next run. [2026-10-09] Verification attempt produced no clear result — recheck next run. Softgen (labelled 'official Model Studio rates') and Requesty list $2.50/$7.50 with cache $0.25; Vercel/Together/OpenRouter resellers list $1.25/$3.75; Merge Gateway $1.65/$4.95. Evidence leans toward list price now applying on Model Studio, but no primary Alibaba read. Prices LEFT at $1.25/$3.75. Third week past the promo expiry — recommend a manual check. [2026-10-10] Input $1.25 → $2.50, Output $3.75 → $7.50 per https://www.alibabacloud.com/en/campaign/qwen-ai-landing-page (Alibaba's own page lists Qwen3.7-Max at $2.5 input / $7.5 output with no promo label; promo has lapsed). Cache Read $0.25 not stated on the page — unchanged, unverified. [2026-10-27] Verification attempt produced no clear result — recheck next run. Snippets surfaced only anotherwrapper $1.25/$3.75 (dated 2026-08-25, i.e. during the promo — stale) and OpenRouter $1.475/$4.425 (reseller). No Alibaba page this run to reconfirm $2.50/$7.50. Prices LEFT at $2.50/$7.50; Cache Read $0.25 still unverified.