Compare · ModelsLive · 2 picked · head to head

StarCoder 2 15B vs Phi 2

Side by side · benchmarks, pricing, and signals you can act on.

Winner summary

StarCoder 2 15B wins 5 of 9 shared benchmarks. Leads in knowledge · language · math.

Category leads
knowledge·StarCoder 2 15Bgeneral·Phi 2language·StarCoder 2 15Bmath·StarCoder 2 15Breasoning·Phi 2
Hype vs Reality
StarCoder 2 15B
#204 by perf·no signal
QUIET
Phi 2
#185 by perf·no signal
QUIET
Best value
StarCoder 2 15B
no price
Phi 2
no price
Vendor risk
U
Unknown
private · undisclosed
Unknown
Microsoft logo
Microsoft
$3.00T·Big Tech
Low risk
Head to head
StarCoder 2 15BPhi 2
ARC AI2
Phi 2 leads by +38.3
AI2 Reasoning Challenge · tests grade-school level science knowledge with multiple-choice questions requiring reasoning beyond simple retrieval.
StarCoder 2 15B
29.6
Phi 2
67.9
BBH (HuggingFace)
Phi 2 leads by +7.7
StarCoder 2 15B
20.4
Phi 2
28.0
GPQA
StarCoder 2 15B leads by +0.2
StarCoder 2 15B
3.1
Phi 2
2.9
IFEval
StarCoder 2 15B leads by +0.4
StarCoder 2 15B
27.8
Phi 2
27.4
MATH Level 5
StarCoder 2 15B leads by +3.0
StarCoder 2 15B
6.0
Phi 2
3.0
MMLU-PRO
Phi 2 leads by +3.1
StarCoder 2 15B
15.0
Phi 2
18.1
MUSR
Phi 2 leads by +10.9
StarCoder 2 15B
2.9
Phi 2
13.8
MMLU
StarCoder 2 15B leads by +7.6
Massive Multitask Language Understanding · 57 subjects spanning STEM, humanities, social sciences, and more. The standard benchmark for broad knowledge.
StarCoder 2 15B
52.1
Phi 2
44.5
Winogrande
StarCoder 2 15B leads by +19.2
WinoGrande · large-scale commonsense reasoning benchmark where models must resolve ambiguous pronouns in carefully constructed sentence pairs.
StarCoder 2 15B
28.6
Phi 2
9.4
Full benchmark table
BenchmarkStarCoder 2 15BPhi 2
ARC AI2
AI2 Reasoning Challenge · tests grade-school level science knowledge with multiple-choice questions requiring reasoning beyond simple retrieval.
29.667.9
BBH (HuggingFace)
20.428.0
GPQA
3.12.9
IFEval
27.827.4
MATH Level 5
6.03.0
MMLU-PRO
15.018.1
MUSR
2.913.8
MMLU
Massive Multitask Language Understanding · 57 subjects spanning STEM, humanities, social sciences, and more. The standard benchmark for broad knowledge.
52.144.5
Winogrande
WinoGrande · large-scale commonsense reasoning benchmark where models must resolve ambiguous pronouns in carefully constructed sentence pairs.
28.69.4
Pricing · per 1M tokens · projected $/mo at 10M tokens
ModelInputOutputContextProjected $/mo
U
StarCoder 2 15B
Microsoft logoPhi 2