Compare · ModelsLive · 2 picked · head to head
Gemma 3 27B (free) vs Llama 3.1 8B Instruct
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Gemma 3 27B (free) wins on 3/3 benchmarks
Gemma 3 27B (free) wins 3 of 3 shared benchmarks. Leads in knowledge · math.
Category leads
knowledge·Gemma 3 27B (free)math·Gemma 3 27B (free)
Hype vs Reality
Attention vs performance
Gemma 3 27B (free)
#172 by perf·#8 by attention
Llama 3.1 8B Instruct
#275 by perf·#18 by attention
Best value
Llama 3.1 8B Instruct
Gemma 3 27B (free)
n/a
$0.00/M
Llama 3.1 8B Instruct
968.0 pts/$
$0.03/M
Vendor risk
Who is behind the model
Google DeepMind
$4.20T·Tier 1
Meta AI
$1.87T·Tier 1
Head to head
3 benchmarks · 2 models
Gemma 3 27B (free)Llama 3.1 8B Instruct
GPQA diamond
Gemma 3 27B (free) leads by +29.2
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Gemma 3 27B (free)
31.8
Llama 3.1 8B Instruct
2.6
MATH level 5
Gemma 3 27B (free) leads by +51.2
MATH Level 5 · the hardest tier of the MATH benchmark, featuring competition-level problems from AMC, AIME, and Olympiad-style mathematics.
Gemma 3 27B (free)
74.0
Llama 3.1 8B Instruct
22.9
OTIS Mock AIME 2024-2025
Gemma 3 27B (free) leads by +18.1
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Gemma 3 27B (free)
19.6
Llama 3.1 8B Instruct
1.6
Full benchmark table
| Benchmark | Gemma 3 27B (free) | Llama 3.1 8B Instruct |
|---|---|---|
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 31.8 | 2.6 |
MATH level 5 MATH Level 5 · the hardest tier of the MATH benchmark, featuring competition-level problems from AMC, AIME, and Olympiad-style mathematics. | 74.0 | 22.9 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 19.6 | 1.6 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.00 | $0.00 | 131K tokens (~66 books) | — | |
| $0.02 | $0.03 | 131K tokens (~66 books) | $0.23 |