Alibaba·LLM — weighted competency·Tue 25 Aug 2026
AQwen3.8 2.4T A95B
Qwen3.8 2.4T A95B scores 92/100 competency for LLM at $3.00.
Moderate confidence
Competency
92/100
category rank #7
Cost
$3.00/1M tok
blended
Speed
24/100
Slow
Accuracy
—
no data yet
Score history.
Weighted LLM competency · last 12 weeks · weekly snapshot.
92/100
+0 pts vs prior
16 Aug19 Aug22 Aug25 Aug
What it's actually good at.
Score on each business-task subset that goes into the weighted competency.
Customer Service20% of category weight
86
Content Writing15% of category weight
88
Research & Analysis16% of category weight
91
Summarization12% of category weight
93
Translation7% of category weight
95
Reasoning16% of category weight
93
Knowledge14% of category weight
72
Pricing.
Published usage pricing. Figures refresh daily from our pricing sources.
$3.00
Usage pricing
$3.00/1M tok
- Input
- $2.00 / 1M tokens
- Output
- $6.00 / 1M tokens
Where it sits.
Position against the rest of the category. Ranked by weighted competency.
| Rank | Model | Score | Cost | Speed | Δ vs Qwen3.8 |
|---|---|---|---|---|---|
| 01 | SGrok 4.6 | 98 | $3.00 | 27 / 100 | +6 |
| 02 | AClaude Opus 5 | 97 | $10.00 | 27 / 100 | +5 |
| 03 | OGPT-5.6 Sol | 96 | $8.00 | 31 / 100 | +4 |
| 04 | AClaude Fable 5 | 95 | $20.00 | 33 / 100 | +3 |
| 05 | GGemini 3.7 Flash | 94 | $1.50 | 60 / 100 | +2 |
| 06 | KKimi K3 | 94 | $6.00 | 18 / 100 | +1 |
| 07 | AQwen3.8 2.4T A95B | 92 | $3.00 | 24 / 100 | — |
| 08 | AQwen3.8 Max | 89 | $3.00 | 17 / 100 | -4 |
| 09 | OGPT-5.5 | 89 | $11.25 | 40 / 100 | -4 |
| 10 | OGPT-5.6 Terra | 88 | $4.50 | 57 / 100 | -4 |
| 11 | DDeepSeek V4 Pro 0813 | 87 | $1.98 | 49 / 100 | -5 |
| 12 | DDeepSeek V4 Flash 0731 | 84 | $0.66 | 78 / 100 | -9 |
| 13 | AClaude Sonnet 5 | 83 | $4.00 | 42 / 100 | -10 |
| 14 | GGemini 3.1 Pro Preview | 81 | $4.50 | 0 / 100 | -12 |
| 15 | ZGLM-5.2 | 80 | $2.15 | 40 / 100 | -12 |