Devstral
Open-weight coding models for autonomous code agents—Devstral 2 and Small 2 with Mistral Vibe CLI.
If you want open-weight coding AI that actually competes with closed models on SWE-bench and costs a fraction, Devstral 2 is the pick. The Vibe CLI makes it immediately useful, and Small 2's local deployability is a real edge. Just know that on human preference, Claude Sonnet 4.5 still wins—so pick Devstral for cost and openness, not for absolute best code quality.
Verified 7d ago · liveness 74/100 · cite: rightaichoice.com/tools/devstral
- Cost-conscious developers seeking open-weight coding AI
- Teams building autonomous coding agents with strong SWE-bench performance
- Enterprises needing on-premise or local model deployment
- Researchers and hobbyists fine-tuning models on consumer hardware
- Non-developers looking for a no-code coding solution
- Users requiring top-tier generation quality that only closed models like Claude Sonnet deliver
- Projects needing context windows beyond 256K
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip Devstral if you prioritize absolute code quality over cost and openness, as Claude Sonnet 4.5 is still preferred in human evaluations.
Devstral 2 is free via API for a limited time; afterward, you'll pay $0.40 per million input tokens and $2.00 per million output tokens, which can add up with high-volume agent workloads.
Devstral's open-weight models are significantly cheaper than closed models like Claude Sonnet—up to 7x more cost-effective on real-world tasks. For API usage, Devstral 2 at $0.40/$2.00 per MTok and Small 2 at $0.10/$0.30 per MTok are budget-friendly. Mistral Vibe plans start at free, with Pro at $14.99/mo, Team at $24.99/user/mo, and Enterprise custom. For cost-conscious teams, this is more affordable than comparable commercial assistants.
In short
Devstral — Open-weight coding models for autonomous code agents—Devstral 2 and Small 2 with Mistral Vibe CLI. Best for Cost-conscious developers seeking open-weight coding AI, Teams building autonomous coding agents with strong SWE-bench performance, Enterprises needing on-premise or local model deployment. Free to start; paid plans from $5.99/mo.
What people actually say about Devstral — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
36 mentions across 2 sources (Hacker News, Lemmy) · researched Jul 3, 2026.
- +Excellent at documentation and writing tasks, better than Claude Opus.
- +SWE-bench Verified score of 72.2% on Devstral 2 (123B).
- +Smaller 24B model runs locally on consumer hardware.
- +Permissive open-source licenses (modified MIT, Apache 2.0).
- +Free API access at launch for experimentation.
- −Mistral is retiring Devstral models in favor of general models.
- −Vibe CLI lags behind Claude Code and other agentic tools.
- −Cost scales exponentially for larger Devstral 2 model.
- −General coding ability is weaker than Claude Sonnet or Opus.
- −Poor at creative tasks like SVG generation.
- • API cost for Devstral 2 (123B) can scale exponentially with usage; not as cheap as claimed.
Viability Score
How well maintained and how widely used is Devstral? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: August 2026
How we score →Key Features
- Open-weight coding models (Devstral 2 and Small 2)
- SWE-bench Verified: 72.2% (Devstral 2), 68.0% (Small 2)
- 256K context window
- Runs locally on consumer GPUs and CPU-only (Small 2)
- Supports image inputs (Small 2)
- Fine-tunable and on-premise deployable
- Mistral Vibe CLI for end-to-end code automation
- Project-aware context scanning
- Multi-file orchestration
- Smart references with @ autocomplete and shell commands
- Tool permission controls
- Agent Communication Protocol integration
- IDE extensions for Zed, VS Code, and JetBrains
- Priced at $0.40/$2.00 per MTok (Devstral 2) after free period
- Voice mode (via Mistral Vibe platform)
About Devstral
Devstral is Mistral AI's family of open-weight coding models designed for end-to-end code automation and as a practical alternative to closed models. Released December 9, 2025, the lineup includes Devstral 2 (123B parameters, modified MIT license) and Devstral Small 2 (24B parameters, Apache 2.0), both engineered for code-agent workflows. With a 256K context window and dense transformer architecture, these models handle complex, multi-file repository tasks while remaining compact enough to run on modest hardware—Small 2 even runs on CPU-only setups. Benchmark performance is strong: Devstral 2 scores 72.2% on SWE-bench Verified, and Small 2 hits 68.0%, outperforming models up to five times its size. Devstral 2 is 5x smaller than DeepSeek V3.2 and 8x smaller than Kimi K2, while matching or exceeding their coding results. That efficiency translates to real cost savings—up to 7x more cost-effective than Claude Sonnet on real-world tasks. Both models support fine-tuning and on-premise deployment. Mistral Vibe CLI is a native, open-source terminal agent powered by Devstral. It provides project-aware context scanning, multi-file orchestration, smart references with @ autocomplete and shell commands, persistent history, and tool permission controls. Vibe CLI integrates with IDEs via Agent Communication Protocol, with extensions for Zed, VS Code, and JetBrains, plus partnerships with Cline and Kilo Code. Devstral 2 is free via API for a limited time; afterward, pricing is $0.40/$2.00 per million tokens (input/output). Small 2 costs $0.10/$0.30 per million tokens. For teams wanting open licenses, local deployment, and tight budgets, Devstral offers a compelling trade-off—though human evals show Claude Sonnet 4.5 still leads on preference.
Behind the Verdict
Here's the thing about Devstral: it's not trying to beat Claude Sonnet on every dimension—it's trying to give you a model you can actually own, run, and afford. If your priority is a coding agent that scores 72.2% on SWE-bench Verified and you don't want to pay per-seat or lock into a closed API, Devstral 2 is a direct hit. We'd reach for it when we need to deploy on-premise or fine-tune on proprietary codebases. The modified MIT and Apache 2.0 licenses mean you can adapt it to your stack without legal friction. The size advantage is real: 5x smaller than DeepSeek V3.2, yet it matches or beats it on coding benchmarks. That translates to lower infrastructure costs, and for startups or enterprises watching GPU budgets, that's the difference between shipping and stalling. Where it bites: human evals still favor Claude Sonnet 4.5 on preference—so if you're optimizing for the absolute best code quality over cost and control, you'll be disappointed. Also, Devstral 2 needs at least 4 H100-class GPUs for deployment, so it's not a 'laptop model.' Small 2 is the one for consumer hardware, including CPU-only. The Vibe CLI is a genuine plus—it's not a wrapper, it's a real terminal agent with project-aware context, multi-file orchestration, and tool permissions. For teams that live in the terminal, that's immediately useful. The integration with Cline and Kilo Code also means you can drop Devstral into your existing agent workflows without rewriting. Compared to alternatives like DeepSeek or Kimi, Devstral's edge is the efficiency-to-performance ratio. You're not paying for a 600B model when a 24B one does the job. That's the kind of pragmatism we like. If you value open licenses, local deployment, and tighter budgets, Devstral is a strong pick—just manage your expectations on
Researching Devstral? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas Devstral actually fits — and what changes day-one when you adopt it.
You're working on a side project and need a coding assistant that doesn't break the bank.
Outcome: Use the free tier of Mistral Vibe, which includes limited coding sessions and API credits, to get started and automate routine refactoring tasks.
Your team wants to automate code reviews and bug fixes without incurring high costs.
Outcome: Adopt Devstral 2 via API in CI/CD pipelines, leveraging its SWE-bench performance to catch issues early, while keeping token costs low.
Your organization requires on-premise AI for compliance reasons.
Outcome: Deploy Devstral Small 2 on your own hardware, fine-tune it on your codebase, and use it for secure, private code assistance.
Use Cases
- Automate bug fixing and feature implementation in large codebases using Mistral Vibe CLI.
- Deploy a local coding assistant on consumer hardware for offline development.
- Integrate Devstral 2 via API into CI/CD pipelines for automated code review and refactoring.
- Fine-tune Devstral Small 2 on proprietary codebases for domain-specific code generation.
- Replace expensive commercial coding assistants with a cost-effective open-weight alternative.
- Run SWE-bench-level coding agents autonomously for software engineering tasks.
- Use Mistral Vibe for long-horizon productivity tasks beyond coding, such as research and analysis.
Models Under the Hood
as of 2026-08-22
Limitations
- The evidence confirms Devstral 2 and Small 2 are open-weight coding models released with Mistral Vibe CLI, designed for autonomous code agents.
- The Small 2 variant can run locally on consumer GPUs and CPU-only, while Devstral 2 offers a larger capability.
- No detailed usage limits or restrictions are specified in the scraped evidence.
as of 2026-08-11
Verification history
We have re-verified Devstral 6 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Plans compared
For each published Devstral tier: who it actually fits, and what it adds vs. the previous tier. Cross-reference the cost calculator above for projected annual outlay.
Free
$0/mo
Ideal for
Casual users and developers exploring Mistral Vibe's capabilities without any cost commitment.
What this tier adds
The entry point offering limited messages, web searches, coding sessions, and image generation, plus $10 in API credits.
Education
$5.99/mo
Ideal for
Verified students at accredited institutions who need more usage than the free tier for academic projects.
What this tier adds
A discounted plan at $5.99/month with more messages and web searches, capped at 12 months.
Pro
$14.99/mo
Ideal for
Professional developers and heavy Vibe users who need all-day coding in CLI, IDE, and web.
What this tier adds
Adds all-day coding, more messages, searches, image generations, and $30 monthly API credits.
Team
$24.99/user/mo
Ideal for
Small to medium teams that need a collaborative workspace with shared context and export capabilities.
What this tier adds
Adds 30GB storage per user, domain name verification, data export, and collaborative workspace features.
Enterprise
Contact sales
Ideal for
Large organizations that require custom models, agents, and workflows, along with security features like SAML SSO.
What this tier adds
Provides custom models, agents, workflows, audit logs, SAML SSO, and white labeling for enterprise-scale needs.
Where the pricing makes sense
The company stage and team size where Devstral's pricing actually pencils out — and where peers do it cheaper.
Devstral's open-weight models are significantly cheaper than closed models like Claude Sonnet—up to 7x more cost-effective on real-world tasks. For API usage, Devstral 2 at $0.40/$2.00 per MTok and Small 2 at $0.10/$0.30 per MTok are budget-friendly. Mistral Vibe plans start at free, with Pro at $14.99/mo, Team at $24.99/user/mo, and Enterprise custom. For cost-conscious teams, this is more affordable than comparable commercial assistants.
Setup time & first value
How long it actually takes to get something useful out of Devstral — broken out by persona, not the marketing-page minute.
For API usage, you can start within minutes by signing up and getting your API key. For local deployment of Small 2, expect 30-60 minutes on a consumer GPU to set up and run. Fine-tuning requires additional time depending on your dataset and hardware.
Switching to or from Devstral
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From GitHub Copilot: Replace Copilot's suggestions with Devstral's open-weight models via API or CLI for more control and lower cost.
- →From ChatGPT Code Interpreter: Use Vibe CLI's project-aware scanning and multi-file orchestration to handle larger codebases.
- ↗To Claude Sonnet: If you need higher quality responses on complex tasks, you can switch your AI backend to Claude while keeping your workflow.
Integrations
Resources & Guides
Tutorials & Learning
Official links
Tools that pair well with Devstral
Common stack mates teams adopt alongside Devstral, with the specific reason each pairing earns its keep.
Poolside AI
Open-weight agentic coding models for regulated enterprises needing auditable, on-prem AI
Zhipu GLM
Chinese enterprise AI platform with open-source GLM models, MaaS APIs, and autonomous agents
OpenHands
Open-source platform for autonomous cloud coding agents that fix bugs, review PRs, and automate workflows.
Featured Head-to-Head Comparisons
Devstral vs Locus Robotics
Choose Locus Robotics if you run a high-volume warehouse needing flexible AMR automation with rapid ROI, backed by RaaS and major WMS integrations. Choose Devstral if you're a developer seeking an open-weight, cost-efficient coding AI with a native CLI agent for autonomous code generation. They serve entirely different domains — logistics vs. software development — so the decision hinges on your operational focus.
Devstral vs Presto Voice
Choose Presto Voice if you run a QSR chain and need proven drive-thru automation with upselling. Choose Devstral if you're a developer or team building autonomous coding pipelines with open-weight models. They serve entirely different domains; your choice depends on whether your need is restaurant operations or software engineering.
Devstral vs Truleo
If you're a law enforcement agency drowning in siloed data and need to surface leads quickly, Truleo is purpose-built for that. If you're a developer seeking a cost-efficient, open-weight coding model with strong SWE-bench performance, Devstral is an excellent choice. They serve entirely different domains—pick based on your role, not a head-to-head feature fight.
Alternatives to Devstral
View allPoolside AI
Open-weight agentic coding models for regulated enterprises needing auditable, on-prem AI
Frequently Asked Questions
Best-of guides
Used Devstral? Help shape our editorial sentiment research.