vs
8:0
overall capability wins (0 ties across 8 scored). Qwen3.8 Max leads overall.
- Reasoning4:0
- Coding1:0
- Pricing0:3
- Speed0:2
- Specs3:0
Verdict
Qwen3.8 Max leads coding (widest gap: +24.3 on SciCode); Qwen3.8 Max offers ~6.3x larger context (1M vs 160K); Qwen3.8 Max allows ~8.0x more max output (131K vs 16K).
- Qwen3.8 Max leads coding (widest gap: +24.3 on SciCode)
- Qwen3.8 Max offers ~6.3x larger context (1M vs 160K)
- Qwen3.8 Max allows ~8.0x more max output (131K vs 16K)
- Qwen3-Coder 30B-A3B is ~25.0x 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-Coder 30B-A3B
- $0.14
Qwen3-Coder 30B-A3B is about 25.5x cheaper on this workload.
Price per 1M tokens
Full comparison
Organization
Alibaba
Alibaba
Family
Qwen
Qwen
License
Qwen3.8-Max License
Apache 2.0
Open weights
Yes
Yes
Release
Aug 3, 2026
Jul 22, 2025
Knowledge cutoff
-
-
API / provider
Alibaba Cloud
Alibaba / Together (ref.)
Modalities
text, image, video → text
text → text
Specs
Context window
1M
160K
Max output
131K
16K
Parameters
2.4T (95B act.)
30B (3B act.)
Pricing
Input $/1M
$2
$0.07
Output $/1M
$6
$0.27
Blended $/1M (3∶1)
$3
$0.12
Speed
tok/s
86
120
TTFT (s)
0.4
0.25
Reasoning
MMLU-Pro
—
70.6
GPQA Diamond
92.8
51.6
Humanity's Last Exam
43.1
3.8
AIME 2025
—
29.0
MATH-500
—
89.3
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
31.7
16.4
AA-LCR v1.1
80.3
32.7
CritPt
17.7
—
MMMU-Pro
82.8
—
IFBench
—
32.7
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
67.7
—
SWE-bench Multilingual
—
—
LiveCodeBench
—
40.3
Terminal-Bench 2.1
81.3
—
Aider Polyglot
—
—
Terminal-Bench 3
—
—
BigCodeBench
—
—
SciCode
52.1
27.8
CursorBench
—
—
SWE-Rebench
—
21.6
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
—
| Metric | Qwen3.8 Max | Qwen3-Coder 30B-A3B | Delta |
|---|---|---|---|
| Identity | |||
| Organization | Alibaba | Alibaba | — |
| Family | Qwen | Qwen | — |
| License | Qwen3.8-Max License | Apache 2.0 | — |
| Open weights | Yes | Yes | — |
| Release | Aug 3, 2026 | Jul 22, 2025 | — |
| Knowledge cutoff | - | - | — |
| API / provider | Alibaba Cloud | Alibaba / Together (ref.) | — |
| Modalities | text, image, video → text | text → text | — |
| Specs | |||
| Context window | 1M | 160K | A +840K |
| Max output | 131K | 16K | A +115K |
| Parameters | 2.4T (95B act.) | 30B (3B act.) | A +2K |
| Pricing | |||
| Input $/1M | $2 | $0.07 | B +$1.93 |
| Output $/1M | $6 | $0.27 | B +$5.73 |
| Blended $/1M (3∶1) | $3 | $0.12 | B +$2.88 |
| Speed | |||
| tok/s | 86 | 120 | B +34 |
| TTFT (s) | 0.4 | 0.25 | B +0.15 |
| Reasoning | |||
| MMLU-Pro | — | 70.6 | — |
| GPQA Diamond | 92.8 | 51.6 | A +41.2 pts |
| Humanity's Last Exam | 43.1 | 3.8 | A +39.3 pts |
| AIME 2025 | — | 29.0 | — |
| MATH-500 | — | 89.3 | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | 31.7 | 16.4 | A +15.3 pts |
| AA-LCR v1.1 | 80.3 | 32.7 | A +47.6 pts |
| CritPt | 17.7 | — | — |
| MMMU-Pro | 82.8 | — | — |
| IFBench | — | 32.7 | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | 67.7 | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | — | 40.3 | — |
| Terminal-Bench 2.1 | 81.3 | — | — |
| Aider Polyglot | — | — | — |
| Terminal-Bench 3 | — | — | — |
| BigCodeBench | — | — | — |
| SciCode | 52.1 | 27.8 | A +24.3 pts |
| CursorBench | — | — | — |
| SWE-Rebench | — | 21.6 | — |
| 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 | — | — |
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-Coder 30B-A3B
- Context
- 160K / 16K out
- Parameters
- 30B (3B act.)
- Price
- $0.07 / $0.27
- Speed
- 120 tok/s · 0.25s TTFT
- Modalities
- text → text
- License
- Apache 2.0
Sparse coding MoE in the Qwen3-Coder line — efficient agent edits without full 480B cost.
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.