LOADING DATASHEET
LOADING DATASHEET
by Mistral
RANKED #14 OF 20 CODE · OVERALL #363 OF 493 RANKED · VIBE SCORE 5.6 · 50 VOICES ·
COPY BADGE pastes a README snippet with the rank SVG.
Mistral coding agent model for repository tasks and software engineering workflows
8 mentions
4 mentions
26 weeks · 26 voices in this window
TOO QUIET
Not enough weekly voices to call a trend yet.
Open weights. You can use a hosted service, or download it and run it yourself, free.
Open weights: download Devstral Small once and it's yours. No subscription, no rate limits, works offline.
24B parameters, a Q4 file is about 14 GB
NVIDIA GeForce RTX 4060 Ti 16GB
16 GB VRAM
~$450 NEWQ4_K_M, ~22 tok/s, 8k context
NVIDIA GeForce RTX 4090 24GB
24 GB VRAM
~$1,750 NEWQ4_K_M to Q5_K_M, ~45 tok/s, 32k context
Apple Mac Studio M2 Ultra (64GB)
64 GB UNIFIED MEMORY
~$3,200 USEDFP16 / Q8_0, ~30 tok/s, full 128k context
The three cards above are researched picks. Search the machine you actually have — we only say yes if it should reply at a usable speed, not just load the weights and crawl.
HARDWARE PICKS & PRICES ARE RESEARCHED FROM THE LIVE WEB AND REFRESHED AUTOMATICALLY. TREAT THEM AS BALLPARK, NOT GOSPEL.
Everyday jobs on each of those rigs: how long Devstral Small takes, and how good the crowd says it is at that kind of work.
| THE JOB | BARE MINIMUMNVIDIA GeForce RTX 4060 Ti 16GB | SWEET SPOTNVIDIA GeForce RTX 4090 24GB | FULL POWERApple Mac Studio M2 Ultra (64GB) |
|---|---|---|---|
| Write an emaila paragraph or two, drafted from a one-line brief | ~23 S★★★★★ | ~11 S★★★★★ | ~17 S★★★★★ |
| Summarize a documenta long report boiled down to the points that matter | ~68 S★★★★★ | ~33 S★★★★★ | ~50 S★★★★★ |
| Build a websitea small landing page, markup and styles together | ~6.1 MIN★★★★★ | ~3 MIN★★★★★ | ~4.4 MIN★★★★★ |
| Build a backendan API with routes, storage and tests | ~19 MIN★★★★★ | ~9.3 MIN★★★★★ | ~14 MIN★★★★★ |
| Build a gamea playable browser game in one file | ~11 MIN★★★★★ | ~5.6 MIN★★★★★ | ~8.3 MIN★★★★★ |
STARS ARE THE COMMUNITY SCORE FOR THAT KIND OF WORK, DOCKED FOR HOW SQUASHED THE WEIGHTS GET AT EACH TIER. TIMES ARE COMPUTED FROM RESEARCHED SPEEDS. BALLPARK, NOT GOSPEL.
Not enough discussion about Devstral Small yet to call it. We found 50 posts but people did not say much either way.
38 POSTS · 12 COMMENTS · REDDIT · LEMMY · GITHUB
Other models the crowd has fully reviewed, starting with code models like this one.
Introducing Devstral Small 24B
Announcing Codestral 25.08 and the complete Mistral Coding Stack for Enterprises
Mistral - Devstral-Small-2507 GGUFs out now!
Introducing Devstral Small 1.1 and Devstral Medium
I benchmarked 21 local LLMs on a MacBook Air M5 for code quality AND speed
Devstral + Vision Dynamic GGUFs out now!
Devstral Small VibeCoded my own Deep Research assistant. Thanks Mistral!
Mistral API quota and rate limits pools analysis for Free Tier plan (20.02.2026)
Magistral Small with Vision
Added Devstral Medium and Small tests to Rival to compare
mistralai/Devstral-Small-2505 · Hugging Face
Any good models for coding (Python and JS) to run on a 16 GB 5080?