About
Devstral Small 1.1 is a 24B parameter open-weight language model for software engineering agents, developed by Mistral AI in collaboration with All Hands AI. Finetuned from Mistral Small 3.1 and...
Related Models
Pricing
Input
$0.10
per 1M tokens
Output
$0.30
per 1M tokens
Blended
$0.15
per 1M tokens
Cheaper than 75% of models. Median price is $0.54/1M tokens.
Cost Calculator
Tokens per day1M
100K100M
Daily
$0.15
Monthly
$4.50
vs. Similar Models
Perplexity: Sonar ProQ:0.0
$6.00+3900%
Qwen3 30B A3B (Reasoning)Q:0.0
$0.18+20%
QwQ 32B-PreviewQ:-0.1
$0.14-10%
Perplexity: SonarQ:+0.2
$1.00+567%
Performance
40
tokens/sec
Faster than 9% of models
0.62
seconds
Faster than 75% of models
0.62
seconds
Faster than 83% of models
Market Median
94 tok/s
57% slower
Median TTFT
1.10s
44% faster
Throughput/Dollar
266
tok/s per $/1M
Speed Comparison
Devstral 2
40 tok/s-0%
Hermes 4 - Llama-3.1 405B (Non-reasoning)
40 tok/s-0%
Qwen3.5 4B (Non-reasoning)
40 tok/s-1%
Context Window
131K
tokens
Larger than 27% of models
Benchmarks
MMLU-Pro
62.2%
GPQA Diamond
41.4%
HLE
3.7%
LiveCodeBench
25.4%
SciCode
24.3%
TerminalBench Hard
6.1%
MATH-500
63.5%
AIME
0.3%
AIME 2025
29.3%
IFBench
34.6%
Long Context Recall
17.0%
Tau2
28.4%
Market AverageTop Score