Devstral
ActiveAgentic coding model for software development.
Updated 4 days agoStructured data from Modeldex catalog
AgenticLong context
Not enough benchmark coverage yet for an Intelligence Index — needs at least 3 results across 2 categories.
History
Devstral is available from Mistral AI.
Training & availability
Mistral AI has not released the underlying model weights — access is via their hosted API only.
Capabilities
-
Context window: 256K tokens.
-
Max output: 256K tokens.
-
Input modalities: text.
Recommended for: agentic, long-context.
Limitations
- Text-only — cannot process images, audio, or video inputs.
Quick start
Minimal example using the OpenRouter API. Copy, paste, replace the key.
from openai import OpenAI
client = OpenAI(
base_url="https://openrouter.ai/api/v1",
api_key="sk-or-...",
)
resp = client.chat.completions.create(
model="mistral/devstral",
messages=[{"role": "user", "content": "Explain quantum computing in one sentence."}],
)
print(resp.choices[0].message.content)Benchmarks
| Benchmark | Score | Source |
|---|---|---|
| SWE-bench VerifiedCoding | 53.6% resolved | Third-party Papers With Code |
Integrations & tooling support
- Tool calling
- Supported
- Structured outputs
- Supported
Price vs quality
Competent benchmarks
Solid benchmark performance. Pricing not publicly available — check the provider.
- Quality percentile
- 50%
- Effective price
- —
- Pricing breakdown
- — in
— out
vs 1 benchmark
pricing not available
Community ratings
No ratings yet. Be the first to rate Devstral.
Rate Devstral
Sign in to rate and review.
Comments
Sign in to leave a comment.