vs
8:0
overall capability wins (0 ties across 8 scored). Qwen3.8 Max leads overall.
- Reasoning4:0
- Coding1:0
- Arena1:0
- Pricing0:3
- Speed2:0
- Specs2:0
Verdict
Qwen3.8 Max leads coding (widest gap: +13.8 on SciCode); Qwen3.8 Max ranks higher on LMArena (+57 Elo); Qwen3.8 Max offers ~3.8x larger context (1M vs 262K).
- Qwen3.8 Max leads coding (widest gap: +13.8 on SciCode)
- Qwen3.8 Max ranks higher on LMArena (+57 Elo)
- Qwen3.8 Max offers ~3.8x larger context (1M vs 262K)
- Qwen3.8 Max allows ~4.0x more max output (131K vs 33K)
- Qwen3.8 Max ships open weights (Qwen3.8-Max License)
- Qwen3 Max is ~1.8x cheaper on a blended token basis than Qwen3.8 Max
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- Qwen3.8 Max
- $3.50
- Qwen3 Max
- $1.75
Qwen3 Max is about 2.0x cheaper on this workload.
Price per 1M tokens
Full comparison
Organization
Alibaba
Alibaba
Family
Qwen
Qwen
License
Qwen3.8-Max License
Proprietary
Open weights
Yes
No
Release
Aug 3, 2026
Sep 1, 2025
Knowledge cutoff
-
-
API / provider
Alibaba Cloud
Alibaba Cloud
Modalities
text, image, video → text
text, image → text
Specs
Context window
1M
262K
Max output
131K
33K
Parameters
2.4T (95B act.)
—
Pricing
Input $/1M
$2
$0.50
Output $/1M
$6
$5
Blended $/1M (3∶1)
$3
$1.63
Speed
tok/s
86
65
TTFT (s)
0.4
0.5
Reasoning
MMLU-Pro
—
84.1
GPQA Diamond
92.8
76.4
Humanity's Last Exam
43.1
11.9
AIME 2025
—
80.7
MATH-500
—
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
31.7
24.4
AA-LCR v1.1
80.3
50.0
CritPt
17.7
—
MMMU-Pro
82.8
—
IFBench
—
44.1
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
67.7
—
SWE-bench Multilingual
—
—
LiveCodeBench
—
76.7
Terminal-Bench 2.1
81.3
—
Aider Polyglot
—
—
Terminal-Bench 3
—
—
BigCodeBench
—
—
SciCode
52.1
38.3
CursorBench
—
—
SWE-Rebench
—
—
NL2Repo-Bench
55.9
—
DeepSWE
56.6
—
WebDev Arena
1,671
—
Terminal-Bench 4.0
—
—
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
1,481
1,424
| Metric | Qwen3.8 Max | Qwen3 Max | Delta |
|---|---|---|---|
| Identity | |||
| Organization | Alibaba | Alibaba | — |
| Family | Qwen | Qwen | — |
| License | Qwen3.8-Max License | Proprietary | — |
| Open weights | Yes | No | — |
| Release | Aug 3, 2026 | Sep 1, 2025 | — |
| Knowledge cutoff | - | - | — |
| API / provider | Alibaba Cloud | Alibaba Cloud | — |
| Modalities | text, image, video → text | text, image → text | — |
| Specs | |||
| Context window | 1M | 262K | A +738K |
| Max output | 131K | 33K | A +98K |
| Parameters | 2.4T (95B act.) | — | — |
| Pricing | |||
| Input $/1M | $2 | $0.50 | B +$1.50 |
| Output $/1M | $6 | $5 | B +$1 |
| Blended $/1M (3∶1) | $3 | $1.63 | B +$1.38 |
| Speed | |||
| tok/s | 86 | 65 | A +21 |
| TTFT (s) | 0.4 | 0.5 | A +0.10 |
| Reasoning | |||
| MMLU-Pro | — | 84.1 | — |
| GPQA Diamond | 92.8 | 76.4 | A +16.4 pts |
| Humanity's Last Exam | 43.1 | 11.9 | A +31.2 pts |
| AIME 2025 | — | 80.7 | — |
| MATH-500 | — | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | 31.7 | 24.4 | A +7.3 pts |
| AA-LCR v1.1 | 80.3 | 50.0 | A +30.3 pts |
| CritPt | 17.7 | — | — |
| MMMU-Pro | 82.8 | — | — |
| IFBench | — | 44.1 | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | 67.7 | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | — | 76.7 | — |
| Terminal-Bench 2.1 | 81.3 | — | — |
| Aider Polyglot | — | — | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | — | — | — |
| SciCode | 52.1 | 38.3 | A +13.8 pts |
| CursorBench | — | — | — |
| SWE-Rebench | — | — | — |
| NL2Repo-Bench | 55.9 | — | — |
| DeepSWE | 56.6 | — | — |
| WebDev Arena | 1,671 | — | — |
| Terminal-Bench 4.0 | — | — | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | 1,481 | 1,424 | A +57 Elo |
Qwen3.8 Max
- Context
- 1M / 131K out
- Parameters
- 2.4T (95B act.)
- Price
- $2 / $6
- Speed
- 86 tok/s · 0.4s TTFT
- Modalities
- text, image, video → text
- License
- Qwen3.8-Max License
Alibaba's Aug 2026 Max flagship (2.4T MoE / 95B active) — frontier agentic coding via QwenCloud / DashScope with 1M multimodal context and $2/$6 API pricing. Open weights released Aug 12, 2026 as Qwen3.8-2.4T-A95B (text-only, thinking-mandatory, Qwen3.8-Max license); the hosted version adds vision input, non-thinking mode, and built-in tools.
Qwen3 Max
- Context
- 262K / 33K out
- Parameters
- —
- Price
- $0.50 / $5
- Speed
- 65 tok/s · 0.5s TTFT
- Modalities
- text, image → text
- License
- Proprietary
Earlier Qwen Max API model; solid coding with competitive Chinese-market pricing.
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.