VerdictPal · editorial desk · updated 1 Sep 2026VerdictPal

Compare · Two models, head-to-head

Pick two models. See the diff.

Composite scores, benchmark ranks, capability grids, and pricing per million tokens. All sourced from the same data that powers the model atlas.

Tools Models

Qwen3.8 2.4T A95B vs Qwen3.8 Max

Model A

Model B

MODEL A

Qwen3.8 2.4T A95B

Alibaba · Qwen

Composite 64Intel 589 rows1 sources

Alibaba's open-weight Qwen-Max-class MoE (12 Aug 2026): 2.4T total / 95B active. Artificial Analysis Intelligence Index 57.7. AA list price $2 / $6 per 1M tokens.

MODEL B

Qwen3.8 Max

Alibaba · Qwen

Composite 66Intel 5810 rows1 sources

Alibaba's August 2026 Max-tier flagship. Artificial Analysis Intelligence Index 58.1. API list price $2 / $6 per 1M tokens on the AA snapshot.

Qwen3.8 2.4T A95B0stats won

Tied3stats

Qwen3.8 Max4stats won

Spec showdown

Capacity, price, freshness.

Qwen3.8 2.4T A95BvsQwen3.8 Max

Composite scores are within 5 points — 64 for Qwen3.8 2.4T A95B vs 66 for Qwen3.8 Max. The choice comes down to capability profile and price. Qwen3.8 Max's 1M context window dwarfs Qwen3.8 2.4T A95B's 262K — a meaningful difference for long-document or codebase work.

Composite score
Qwen3.8 2.4T A95B64Qwen3.8 Max66Winner on this row
Intelligence (Vals/AA)
Qwen3.8 2.4T A95B58Qwen3.8 Max58
Context window
Qwen3.8 2.4T A95B262KQwen3.8 Max1MWinner on this row
Max output
Qwen3.8 2.4T A95B—Qwen3.8 Max—
Input price / 1M (lower wins)
Qwen3.8 2.4T A95B$2.0Qwen3.8 Max$2.0
Output speed (AA default API)
Qwen3.8 2.4T A95B40 t/sQwen3.8 Max40 t/sWinner on this row
Time to first token (AA default API)
Qwen3.8 2.4T A95B1.76sQwen3.8 Max1.46sWinner on this row

Composite score

Intelligence (Vals/AA)

Context window

262K

Max output

—

Input price / 1M (lower wins)

$2.0

Output speed (AA default API)

40 t/s

Time to first token (AA default API)

1.76s

1.46s

Capability grid

What each model can do.

Modalities

Qwen3.8 2.4T A95BQwen3.8 Max

Text●●

Image—●

Audio——

Video——

Code●●

Tool use●●

Reasoning●●

Access

Qwen3.8 2.4T A95BQwen3.8 Max

API●●

Consumer app——

Open weights●—

Managed cloud—●

On-device——

Family profile

Top score per benchmark family.

Reasoning

Coding

Agentic

Long context

Benchmark ranks

Every shared benchmark, side by side.

Reasoning

Artificial Analysis Intelligence Index#12/184 · #11/184

57.7%

58.1%

GPQA Diamond#9/185 · #16/185

93.5%

92.7%

Humanity's Last Exam#24/180 · #17/180

42.4%

43%

Coding

SciCode#34/180 · #28/180

51.6%

52.9%

Agentic

Terminal-Bench#16/184 · #20/184

82.02%

81.27%

τ³-Bench Banking#4/106 · #1/106

49.07%

51.34%

Long context

AA-LCR#47/179 · #56/179

75.33%

74.33%

Performance

AA output speed#78/185 · #77/185

40 t/s

AA time to first token#33/185 · #45/185

1.76s

1.46s

Dossier facts

Dates, sources, fineprint.

Qwen3.8 2.4T A95BQwen3.8 Max

ProviderAlibabaAlibaba

FamilyQwenQwen

Release date2026-08-122026-08-03

Knowledge cutoffunpublishedunpublished

Context window262K1M

Max output——

Input price /1M$2.0$2.0

Output price /1M$6.0$6.0

Statussolidflagship

Composite score6466

Intelligence score5858

Intel sourceaaaa

Benchmark rows910

Source count11

Freshnessfreshfresh

Pricing checked2026-08-152026-08-12

Ecosystem

Which tools wrap each model.

Qwen3.8 2.4T A95B

Not listed in any tool yet.

Qwen3.8 Max

Not listed in any tool yet.

Verdict

Which model to actually pick.

Open Qwen3.8 2.4T A95B's dossier Open Qwen3.8 Max's dossier