Compare · ModelsLive · 2 picked · head to head
Yi 6B vs Llama 2-13B
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Llama 2-13B wins on 4/7 benchmarks
Llama 2-13B wins 4 of 7 shared benchmarks. Leads in knowledge · reasoning.
Category leads
knowledge·Llama 2-13Breasoning·Llama 2-13Bmath·Yi 6B
Hype vs Reality
Attention vs performance
Yi 6B
#183 by perf·no signal
Llama 2-13B
#128 by perf·no signal
Vendor risk
Who is behind the model
U
Unknown
private · undisclosed
Meta AI
$1.50T·Tier 1
Head to head
7 benchmarks · 2 models
Yi 6BLlama 2-13B
ARC AI2
Llama 2-13B leads by +13.3
AI2 Reasoning Challenge · tests grade-school level science knowledge with multiple-choice questions requiring reasoning beyond simple retrieval.
Yi 6B
33.7
Llama 2-13B
47.1
BBH
Llama 2-13B leads by +14.7
BIG-Bench Hard · a curated subset of 23 challenging tasks from BIG-Bench where language models previously failed to outperform average humans.
Yi 6B
29.6
Llama 2-13B
44.3
GSM8K
Yi 6B leads by +8.0
Grade School Math 8K · 8,500 linguistically diverse grade-school math word problems that require multi-step reasoning to solve.
Yi 6B
44.9
Llama 2-13B
36.9
HellaSwag
Llama 2-13B leads by +8.4
HellaSwag · tests commonsense reasoning by asking models to predict the most plausible continuation of everyday scenarios.
Yi 6B
65.9
Llama 2-13B
74.3
MATH level 5
Yi 6B leads by +1.9
MATH Level 5 · the hardest tier of the MATH benchmark, featuring competition-level problems from AMC, AIME, and Olympiad-style mathematics.
Yi 6B
5.2
Llama 2-13B
3.3
MMLU
Yi 6B leads by +11.2
Massive Multitask Language Understanding · 57 subjects spanning STEM, humanities, social sciences, and more. The standard benchmark for broad knowledge.
Yi 6B
52.0
Llama 2-13B
40.8
Winogrande
Llama 2-13B leads by +3.0
WinoGrande · large-scale commonsense reasoning benchmark where models must resolve ambiguous pronouns in carefully constructed sentence pairs.
Yi 6B
42.6
Llama 2-13B
45.6
Full benchmark table
| Benchmark | Yi 6B | Llama 2-13B |
|---|---|---|
ARC AI2 AI2 Reasoning Challenge · tests grade-school level science knowledge with multiple-choice questions requiring reasoning beyond simple retrieval. | 33.7 | 47.1 |
BBH BIG-Bench Hard · a curated subset of 23 challenging tasks from BIG-Bench where language models previously failed to outperform average humans. | 29.6 | 44.3 |
GSM8K Grade School Math 8K · 8,500 linguistically diverse grade-school math word problems that require multi-step reasoning to solve. | 44.9 | 36.9 |
HellaSwag HellaSwag · tests commonsense reasoning by asking models to predict the most plausible continuation of everyday scenarios. | 65.9 | 74.3 |
MATH level 5 MATH Level 5 · the hardest tier of the MATH benchmark, featuring competition-level problems from AMC, AIME, and Olympiad-style mathematics. | 5.2 | 3.3 |
MMLU Massive Multitask Language Understanding · 57 subjects spanning STEM, humanities, social sciences, and more. The standard benchmark for broad knowledge. | 52.0 | 40.8 |
Winogrande WinoGrande · large-scale commonsense reasoning benchmark where models must resolve ambiguous pronouns in carefully constructed sentence pairs. | 42.6 | 45.6 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
U Yi 6B | — | — | — | — |
| — | — | — | — |