Compare · ModelsLive · 2 picked · head to head
Mistral Large 4 vs Mistral Medium 3
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Mistral Large 4 wins on 2/2 benchmarks
Mistral Large 4 wins 2 of 2 shared benchmarks. Leads in knowledge · math.
Category leads
knowledge·Mistral Large 4math·Mistral Large 4
Hype vs Reality
Attention vs performance
Mistral Large 4
#224 by perf·#14 by attention
Mistral Medium 3
#199 by perf·no signal
Best value
Mistral Medium 3
1.3x better value than Mistral Large 4
Mistral Large 4
26.1 pts/$
$1.39/M
Mistral Medium 3
32.8 pts/$
$1.20/M
Vendor risk
Who is behind the model
Mistral AI
$14.0B·Tier 1
Mistral AI
$14.0B·Tier 1
Head to head
2 benchmarks · 2 models
Mistral Large 4Mistral Medium 3
GPQA diamond
Mistral Large 4 leads by +38.1
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Mistral Large 4
84.1
Mistral Medium 3
46.0
OTIS Mock AIME 2024-2025
Mistral Large 4 leads by +66.2
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Mistral Large 4
98.3
Mistral Medium 3
32.1
Full benchmark table
| Benchmark | Mistral Large 4 | Mistral Medium 3 |
|---|---|---|
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 84.1 | 46.0 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 98.3 | 32.1 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.68 | $2.09 | 1.0M tokens (~524 books) | $10.32 | |
| $0.40 | $2.00 | 131K tokens (~66 books) | $8.00 |
People also compared