LimitDigest
Home / Compare / AI models / Qwen4-Max vs Qwen3.8 Max

Qwen4-Max vs Qwen3.8 Max

Official Qwen API prices and LMArena scores, checked 2026-11-15. Add up to 2 more models to compare 4.

Verdict

Qwen4-Max is cheaper and wins more LMArena tasks.

Qwen4-Max is 1.3x cheaper on input and output. Qwen4-Max scores 28 points higher for coding. Across 2 shared LMArena tasks, Qwen4-Max leads 2 to 0. A medium month (20M input + 5M output tokens) costs $52.50 on Qwen4-Max and $70 on Qwen3.8 Max.

  • Pick Qwen4-Max if you want the lower bill ($52.50 vs $70 for a medium month) or you care most about coding (+28) and agent tasks (+23).
  • Pick Qwen3.8 Max if nothing here clearly favors it (no lower bill, bigger context, cheaper cache or LMArena lead of 3+ points over Qwen4-Max).
Sample month

Medium: 20M input + 5M output tokens a month at list price (short-context rate, no caching or batch discounts).

Qwen4-MaxvsQwen3.8 Max
Price per 1M tokens
Input$1.50 (better)$2
Cached input$0.075n/a
Output$4.50 (better)$6
Sample month, medium$52.50 (better)$70
Context
Context window1M1M
Quality (LMArena relative score, out of 100)
Coding72 (better)44
Chatn/a24
Creative writingn/a33
Roleplay and charactersn/a21
Reasoning and mathn/a52
Visionn/a100
Long documentsn/an/a
Cheap bulk and agents40 (better)17

Marked side is better: lower for prices, higher for context and scores. Bars are scaled to the larger of the two.

Spec
Qwen4-Max
Qwen3.8 Max
Price per 1M tokens
Input$1.50$2
Cached input$0.075None
Output$4.50$6
Blend 3:1$2.25$3
Sample month, medium$52.50$70
Context
Context window1M tokens1M tokens
Quality (LMArena, out of 100)
Coding72/100 #6 of 5444/100 #14 of 54
ChatNot ranked24/100 #14 of 54
Creative writingNot ranked33/100 #10 of 54
Roleplay and charactersNot ranked21/100 #8 of 54
Reasoning and mathNot ranked52/100 #7 of 52
VisionNot ranked100/100 #1 of 33
Long documentsNot rankedNot ranked
Cheap bulk and agents40/100 #11 of 4217/100 #30 of 42
Details
ProviderQwenQwen
API model idqwen4-maxqwen3.8-max
NotesInternational (Singapore), launched Nov 16, 2026. List price for all prompts up to 1M tokens. Cache read $0.075; 5-minute cache write $1.875, 1-hour $3. Closed weights.International (Singapore). Alibaba lists the cache-hit price only in the Model Studio console.
Sources
Last checked2026-11-162026-11-15
Official pricingQwen pricingQwen pricing

API prices in USD per 1M tokens from each provider's official pricing page. Blend 3:1 = (3 x input + output) / 4. Sample month = stated input and output tokens at the list price, no caching. Quality: relative score out of 100 on each LMArena leaderboard (100 = top model we price on that board, n/a = not ranked there), data licensed CC BY 4.0. See how we pick.

Where each one ranks on LMArena

Relative score out of 100 on each leaderboard (100 = the top model we price there), with the rank among the models from our API price table that appear on that board.

  • Coding (Code Arena): Qwen4-Max #6 of 54 (72/100), Qwen3.8 Max #14 of 54 (44/100)
  • Chat (Text Arena): Qwen4-Max not ranked, Qwen3.8 Max #14 of 54 (24/100)
  • Creative writing (Text Arena creative writing): Qwen4-Max not ranked, Qwen3.8 Max #10 of 54 (33/100)
  • Roleplay and characters (Text Arena multi-turn): Qwen4-Max not ranked, Qwen3.8 Max #8 of 54 (21/100)
  • Reasoning and math (Text Arena math): Qwen4-Max not ranked, Qwen3.8 Max #7 of 52 (52/100)
  • Vision (Vision Arena): Qwen4-Max not ranked, Qwen3.8 Max #1 of 33 (100/100)
  • Long documents (Document Arena): Qwen4-Max not ranked, Qwen3.8 Max not ranked
  • Cheap bulk and agents (Agent Arena): Qwen4-Max #11 of 42 (40/100), Qwen3.8 Max #30 of 42 (17/100)

Pricing details from the official pages

Qwen4-Max lists $1.50 input, $4.50 output, $0.075 cached input per 1M tokens on Qwen pricing. Qwen3.8 Max lists $2 input, $6 output per 1M tokens on Qwen pricing. Prices checked 2026-11-15.

  • Qwen4-Max: International (Singapore), launched Nov 16, 2026. List price for all prompts up to 1M tokens. Cache read $0.075; 5-minute cache write $1.875, 1-hour $3. Closed weights.
  • Qwen3.8 Max: International (Singapore). Alibaba lists the cache-hit price only in the Model Studio console.

Sample monthly bill at list price, no caching: Light (2M input + 0.5M output tokens): $5.25 vs $7. Medium (20M input + 5M output tokens): $52.50 vs $70. Heavy (200M input + 50M output tokens): $525 vs $700.

FAQ

Is Qwen4-Max cheaper than Qwen3.8 Max?

Qwen4-Max costs $1.50 input and $4.50 output per 1M tokens; Qwen3.8 Max costs $2 and $6. For a medium month (20M input + 5M output tokens) Qwen4-Max is cheaper: $52.50 vs $70.

Which is better for coding, Qwen4-Max or Qwen3.8 Max?

On LMArena's Code Arena Qwen4-Max scores 72/100 (#6 of 54) and Qwen3.8 Max 44/100 (#14 of 54), so Qwen4-Max is ahead by 28 points.

Which has the bigger context window?

Qwen4-Max: 1M tokens. Qwen3.8 Max: 1M tokens. Both list the same window.

How much do Qwen4-Max and Qwen3.8 Max cost per month?

Light (2M input + 0.5M output tokens): $5.25 vs $7. Medium (20M input + 5M output tokens): $52.50 vs $70. Heavy (200M input + 50M output tokens): $525 vs $700. Pick your own token counts in the plan vs API calculator.

Where do these numbers come from?

Prices come from each provider's official pricing page, checked daily. Quality is the relative score on the LMArena leaderboard dataset, licensed CC BY 4.0. Where a model is not ranked we show n/a.

Sources

Related