Compare · ModelsLive · 2 picked · head to head
Stable Beluga 2 vs Qwen-14B
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Stable Beluga 2 wins on 6/6 benchmarks
Stable Beluga 2 wins 6 of 6 shared benchmarks. Leads in knowledge · reasoning · math.
Category leads
knowledge·Stable Beluga 2reasoning·Stable Beluga 2math·Stable Beluga 2
Hype vs Reality
Attention vs performance
Stable Beluga 2
#102 by perf·no signal
Qwen-14B
#37 by perf·no signal
Vendor risk
Who is behind the model
U
Unknown
private · undisclosed
Alibaba (Qwen)
$293.0B·Tier 1
Head to head
6 benchmarks · 2 models
Stable Beluga 2Qwen-14B
ARC AI2
Stable Beluga 2 leads by +2.3
AI2 Reasoning Challenge · tests grade-school level science knowledge with multiple-choice questions requiring reasoning beyond simple retrieval.
Stable Beluga 2
81.5
Qwen-14B
79.2
BBH
Stable Beluga 2 leads by +19.1
BIG-Bench Hard · a curated subset of 23 challenging tasks from BIG-Bench where language models previously failed to outperform average humans.
Stable Beluga 2
59.1
Qwen-14B
40.0
GSM8K
Stable Beluga 2 leads by +8.3
Grade School Math 8K · 8,500 linguistically diverse grade-school math word problems that require multi-step reasoning to solve.
Stable Beluga 2
69.6
Qwen-14B
61.3
LAMBADA
Stable Beluga 2 leads by +0.2
LAMBADA · measures the ability to predict the final word of a passage, requiring broad contextual understanding across long text spans.
Stable Beluga 2
71.3
Qwen-14B
71.1
MMLU
Stable Beluga 2 leads by +3.1
Massive Multitask Language Understanding · 57 subjects spanning STEM, humanities, social sciences, and more. The standard benchmark for broad knowledge.
Stable Beluga 2
58.1
Qwen-14B
55.1
PIQA
Stable Beluga 2 leads by +6.8
PIQA (Physical Interaction QA) · tests intuitive physical reasoning by asking models to select the correct approach for everyday physical tasks.
Stable Beluga 2
66.6
Qwen-14B
59.8
Full benchmark table
| Benchmark | Stable Beluga 2 | Qwen-14B |
|---|---|---|
ARC AI2 AI2 Reasoning Challenge · tests grade-school level science knowledge with multiple-choice questions requiring reasoning beyond simple retrieval. | 81.5 | 79.2 |
BBH BIG-Bench Hard · a curated subset of 23 challenging tasks from BIG-Bench where language models previously failed to outperform average humans. | 59.1 | 40.0 |
GSM8K Grade School Math 8K · 8,500 linguistically diverse grade-school math word problems that require multi-step reasoning to solve. | 69.6 | 61.3 |
LAMBADA LAMBADA · measures the ability to predict the final word of a passage, requiring broad contextual understanding across long text spans. | 71.3 | 71.1 |
MMLU Massive Multitask Language Understanding · 57 subjects spanning STEM, humanities, social sciences, and more. The standard benchmark for broad knowledge. | 58.1 | 55.1 |
PIQA PIQA (Physical Interaction QA) · tests intuitive physical reasoning by asking models to select the correct approach for everyday physical tasks. | 66.6 | 59.8 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
U Stable Beluga 2 | — | — | — | — |
| — | — | — | — |