Price Per TokenPrice Per Token
Meta-llama
Meta-llama
vs
Mistral AI
Mistral AI

Llama 3.1 405B Instruct vs Devstral 2 2512

A detailed comparison of pricing, benchmarks, and capabilities

108 out of our 483 tracked models have had a price change in March.

Get our weekly newsletter on pricing changes, new releases, and tools.

Key Takeaways

Llama 3.1 405B Instruct wins:

  • No clear advantages in compared metrics

Devstral 2 2512 wins:

  • Cheaper input tokens
  • Larger context window
  • Faster response time
  • Higher intelligence benchmark
  • Better at coding
  • Better at math
Price Advantage
Devstral 2 2512
Benchmark Advantage
Devstral 2 2512
Context Window
Devstral 2 2512
Speed
Devstral 2 2512

Pricing Comparison

Price Comparison

MetricLlama 3.1 405B InstructDevstral 2 2512Winner
Input (per 1M tokens)$0.90$0.40 Devstral 2 2512
Output (per 1M tokens)$0.90$0.90Tie
Cache Read (per 1M)$0.45$0.45Tie
Using a 3:1 input/output ratio, Devstral 2 2512 is 42% cheaper overall.

Llama 3.1 405B Instruct Providers

No provider data available

Devstral 2 2512 Providers

No provider data available

Benchmark Comparison

8
Benchmarks Compared
0
Llama 3.1 405B Instruct Wins
6
Devstral 2 2512 Wins

Benchmark Scores

BenchmarkLlama 3.1 405B InstructDevstral 2 2512Winner
Intelligence Index
Overall intelligence score
17.422.0
Coding Index
Code generation & understanding
14.523.7
Math Index
Mathematical reasoning
3.036.7
MMLU Pro
Academic knowledge
73.276.2
GPQA
Graduate-level science
51.559.4
LiveCodeBench
Competitive programming
30.544.8
Aider
Real-world code editing
66.2--
AIME
Competition math
21.3--
Devstral 2 2512 significantly outperforms in coding benchmarks.

Cost vs Quality

X-axis:
Y-axis:
Loading chart...
Llama 3.1 405B Instruct
Other models

Context & Performance

Context Window

Llama 3.1 405B Instruct
131,000
tokens
Devstral 2 2512
262,144
tokens
Devstral 2 2512 has a 50% larger context window.

Speed Performance

MetricLlama 3.1 405B InstructDevstral 2 2512Winner
Tokens/second33.7 tok/s81.2 tok/s
Time to First Token0.71s0.39s
Devstral 2 2512 responds 141% faster on average.

Capabilities

Feature Comparison

FeatureLlama 3.1 405B InstructDevstral 2 2512
Vision (Image Input)
Tool/Function Calls
Reasoning Mode
Audio Input
Audio Output
PDF Input
Prompt Caching
Web Search

License & Release

PropertyLlama 3.1 405B InstructDevstral 2 2512
LicenseOpen SourceOpen Source
AuthorMeta-llamaMistral AI
ReleasedJul 2024Dec 2025

Llama 3.1 405B Instruct Modalities

Input
text
Output
text

Devstral 2 2512 Modalities

Input
text
Output
text

Related Comparisons

Compare Llama 3.1 405B Instruct with:

Compare Devstral 2 2512 with:

Frequently Asked Questions

Devstral 2 2512 has cheaper input pricing at $0.40/M tokens. Llama 3.1 405B Instruct has cheaper output pricing at $0.90/M tokens.
Devstral 2 2512 scores higher on coding benchmarks with a score of 23.7, compared to Llama 3.1 405B Instruct's score of 14.5.
Llama 3.1 405B Instruct has a 131,000 token context window, while Devstral 2 2512 has a 262,144 token context window.
Llama 3.1 405B Instruct does not support vision. Devstral 2 2512 does not support vision.