Compare · ModelsLive · 2 picked · head to head
StarCoder 2 15B vs LLaMA-13B
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
LLaMA-13B wins on 5/10 benchmarks
LLaMA-13B wins 5 of 10 shared benchmarks. Leads in knowledge · general.
Category leads
knowledge·LLaMA-13Bmath·StarCoder 2 15Bgeneral·LLaMA-13Blanguage·StarCoder 2 15Breasoning·StarCoder 2 15B
Hype vs Reality
Attention vs performance
StarCoder 2 15B
#204 by perf·no signal
LLaMA-13B
#170 by perf·no signal
Vendor risk
Who is behind the model
U
Unknown
private · undisclosed
Meta AI
$1.50T·Tier 1
Head to head
10 benchmarks · 2 models
StarCoder 2 15BLLaMA-13B
ARC AI2
LLaMA-13B leads by +7.3
AI2 Reasoning Challenge · tests grade-school level science knowledge with multiple-choice questions requiring reasoning beyond simple retrieval.
StarCoder 2 15B
29.6
LLaMA-13B
36.9
GSM8K
StarCoder 2 15B leads by +37.2
Grade School Math 8K · 8,500 linguistically diverse grade-school math word problems that require multi-step reasoning to solve.
StarCoder 2 15B
57.7
LLaMA-13B
20.6
BBH (HuggingFace)
LLaMA-13B leads by +4.9
StarCoder 2 15B
20.4
LLaMA-13B
25.3
GPQA
LLaMA-13B leads by +0.3
StarCoder 2 15B
3.1
LLaMA-13B
3.5
IFEval
StarCoder 2 15B leads by +2.5
StarCoder 2 15B
27.8
LLaMA-13B
25.3
MATH Level 5
StarCoder 2 15B leads by +2.9
StarCoder 2 15B
6.0
LLaMA-13B
3.1
MMLU-PRO
LLaMA-13B leads by +8.0
StarCoder 2 15B
15.0
LLaMA-13B
23.1
MUSR
StarCoder 2 15B leads by +1.0
StarCoder 2 15B
2.9
LLaMA-13B
2.0
MMLU
StarCoder 2 15B leads by +21.9
Massive Multitask Language Understanding · 57 subjects spanning STEM, humanities, social sciences, and more. The standard benchmark for broad knowledge.
StarCoder 2 15B
52.1
LLaMA-13B
30.3
Winogrande
LLaMA-13B leads by +17.4
WinoGrande · large-scale commonsense reasoning benchmark where models must resolve ambiguous pronouns in carefully constructed sentence pairs.
StarCoder 2 15B
28.6
LLaMA-13B
46.0
Full benchmark table
| Benchmark | StarCoder 2 15B | LLaMA-13B |
|---|---|---|
ARC AI2 AI2 Reasoning Challenge · tests grade-school level science knowledge with multiple-choice questions requiring reasoning beyond simple retrieval. | 29.6 | 36.9 |
GSM8K Grade School Math 8K · 8,500 linguistically diverse grade-school math word problems that require multi-step reasoning to solve. | 57.7 | 20.6 |
BBH (HuggingFace) | 20.4 | 25.3 |
GPQA | 3.1 | 3.5 |
IFEval | 27.8 | 25.3 |
MATH Level 5 | 6.0 | 3.1 |
MMLU-PRO | 15.0 | 23.1 |
MUSR | 2.9 | 2.0 |
MMLU Massive Multitask Language Understanding · 57 subjects spanning STEM, humanities, social sciences, and more. The standard benchmark for broad knowledge. | 52.1 | 30.3 |
Winogrande WinoGrande · large-scale commonsense reasoning benchmark where models must resolve ambiguous pronouns in carefully constructed sentence pairs. | 28.6 | 46.0 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
U StarCoder 2 15B | — | — | — | — |
| — | — | — | — |