Key Takeaways
Llama 3.2 3B Instruct wins:
- Cheaper output tokens
- Faster response time
- Better at math
Mistral Nemo wins:
- Higher intelligence benchmark
Price Advantage
Llama 3.2 3B Instruct
Benchmark Advantage
Llama 3.2 3B Instruct
Context Window
Mistral Nemo
Speed
Llama 3.2 3B Instruct
Pricing Comparison
Price Comparison
| Metric | Llama 3.2 3B Instruct | Mistral Nemo | Winner |
|---|---|---|---|
| Input (per 1M tokens) | $0.02 | $0.02 | Tie |
| Output (per 1M tokens) | $0.02 | $0.04 | Llama 3.2 3B Instruct |
Using a 3:1 input/output ratio, Llama 3.2 3B Instruct is 20% cheaper overall.
Llama 3.2 3B Instruct Providers
DeepInfra $0.02 (Cheapest)
Novita $0.03
Cloudflare $0.05
Together $0.06
Hyperbolic $0.10
Mistral Nemo Providers
Chutes $0.02 (Cheapest)
DeepInfra $0.02 (Cheapest)
Novita $0.04
Mistral $0.15
Azure $0.30
Benchmark Comparison
7
Benchmarks Compared
2
Llama 3.2 3B Instruct Wins
1
Mistral Nemo Wins
Benchmark Scores
| Benchmark | Llama 3.2 3B Instruct | Mistral Nemo | Winner |
|---|---|---|---|
Intelligence Index Overall intelligence score | 9.7 | 25.0 | |
Math Index Mathematical reasoning | 3.3 | - | - |
MMLU Pro Academic knowledge | 34.7 | 26.4 | |
GPQA Graduate-level science | 25.5 | 8.7 | |
LiveCodeBench Competitive programming | 8.3 | - | - |
AIME Competition math | 6.7 | - | - |
BBH Big-Bench Hard | - | 36.1 | - |
Llama 3.2 3B Instruct wins 2 out of 7 benchmarks.
Cost vs Quality
X-axis:
Y-axis:
Loading chart...
Llama 3.2 3B Instruct
Other models
Context & Performance
Context Window
Llama 3.2 3B Instruct
131,072
tokens
Max output: 16,384 tokens
Mistral Nemo
131,072
tokens
Max output: 16,384 tokens
Speed Performance
| Metric | Llama 3.2 3B Instruct | Mistral Nemo | Winner |
|---|---|---|---|
| Tokens/second | 44.1 tok/s | N/A | |
| Time to First Token | 0.43s | N/A |
Capabilities
Feature Comparison
| Feature | Llama 3.2 3B Instruct | Mistral Nemo |
|---|---|---|
| Vision (Image Input) | ||
| Tool/Function Calls | ||
| Reasoning Mode | ||
| Audio Input | ||
| Audio Output | ||
| PDF Input | ||
| Prompt Caching | ||
| Web Search |
License & Release
| Property | Llama 3.2 3B Instruct | Mistral Nemo |
|---|---|---|
| License | Open Source | Open Source |
| Author | Meta-llama | Mistral AI |
| Released | Sep 2024 | Jul 2024 |
Llama 3.2 3B Instruct Modalities
Input
text
Output
text
Mistral Nemo Modalities
Input
text
Output
text