Devstral Medium
Intelligence
18.7
#275/521
Coding
15.9
#250/427
Math
4.7
#248/265
Speed
136 tok/s
TTFT: 412.00s
Pricing
$0.40 / $2.00
per 1M tokens (in/out)
Devstral Medium is Mistral’s offering for general-purpose AI tasks. It processes at 136.083 tokens per second and is priced at $0.4 per million input tokens, targeting professional users.
When to Use Devstral Medium
✓ Best For
- Natural language processing tasks.
- Data analysis and reporting.
- Coding assistance and debugging.
✗ Not Ideal For
- High-complexity mathematical computations.
- Real-time applications requiring very low latency.
How Devstral Medium Compares
Intelligence Index · Higher is better
MBZUAI Institute of Foundation ModelsAnthropicMistralNous Research
Benchmark Profile
Coding Index
Nous ResearchDeepSeekMistralNVIDIAMeta
Output Speed · tok/s
AnthropicMistralZ AIXiaomi
Math Index
AmazonIBMMistralMeta
Intelligence · Coding · Math
Intelligence
Coding
Math
All Benchmark Scores (16)
| Benchmark | Score | |
|---|---|---|
| Intelligence Index | 18.7 | |
| Coding Index | 15.9 | |
| Math Index | 4.7 | |
| MMLU-Pro | 708% | |
| GPQA | 492% | |
| LiveCodeBench | 337% | |
| HLE | 38% | |
| SWE-Bench Verified | 61.6% | |
| SciCode | 29.4% | |
| IFBench | 29.9% | |
| LCR | 28.7% | |
| TerminalBench Hard | 9.1% | |
| Tau2 | 19.9% | |
| AIME | 6.7% | |
| AIME 2025 | 4.7% | |
| MATH 500 | 70.7% |
Data: Artificial Analysis · Updated: April 2, 2026
Frequently Asked Questions (15)
When was Devstral Medium released?
Devstral Medium was released on July 10, 2025.
Who created Devstral Medium?
Devstral Medium was created by Mistral.
How intelligent is Devstral Medium?
Devstral Medium scores 19 on the Artificial Analysis Intelligence Index, placing it above average among other non-reasoning models in a similar price tier (median: 15).
How fast is Devstral Medium?
Devstral Medium generates output at 129.3 tokens per second (based on Mistral's API), which is above average compared to other non-reasoning models in a similar price tier (median: 91.4 t/s).
What is the latency of Devstral Medium?
Devstral Medium has a time to first token (TTFT) of 1.21s (based on Mistral's API), which is better than average compared to other non-reasoning models in a similar price tier (median: 1.38s).
How much does Devstral Medium cost?
Devstral Medium costs $0.40 per 1M input tokens (at the higher end, median: $0.20) and $2.00 per 1M output tokens (at the higher end, median: $0.70), based on Mistral's API.
What is Devstral Medium API pricing?
Devstral Medium costs $0.40 per 1M input tokens and $2.00 per 1M output tokens (based on Mistral's API). For a blended rate (3:1 input to output ratio), this is $0.80 per 1M tokens. Pricing may vary by provider.
How verbose is Devstral Medium?
When evaluated on the Intelligence Index, Devstral Medium generated 3.4M output tokens, which is very competitive compared to other non-reasoning models in a similar price tier (median: 5.8M).
Is Devstral Medium a reasoning model?
No, Devstral Medium is not a reasoning model. It provides direct responses without extended chain-of-thought reasoning.
What input modalities does Devstral Medium support?
Devstral Medium supports text input.
What output modalities does Devstral Medium support?
Devstral Medium supports text output.
Can Devstral Medium process images?
No, Devstral Medium does not support image input. It can only process text.
Is Devstral Medium multimodal?
No, Devstral Medium is not multimodal. It only supports text input.
What is the context window of Devstral Medium?
Devstral Medium has a context window of 260k tokens. This determines how much text and conversation history the model can process in a single request.
Is Devstral Medium open source?
No, Devstral Medium is proprietary. The model weights are not publicly available.
