Devstral Small 2507 devstral-small-2507

mistralai · mistral coding

#53 overall #20 agentic

params
24B
arch
dense
license
apache-2.0 open weights
released
Jul 2025
downloads/30d
63.0k

# architecture

attn shared GQA 32q/8kv feed-forward (dense) every parameter, every token out ×40 layers 24B — no routing
layers
40
d_model
5120
heads
32q / 8kv
head dim
128
vocab
131072
family
mistral

signal path of one layer, generated from the registry's structured fields — dimension lines quote the model's real numbers; an MoE trace forks at the router, a dense trace runs straight through. Geometry fields come from the repo's config.json.

# why ranked

Overall: #53 — score 15.7 ● high — all signals present

signalweightinput (0–100)
aa_intelligence 0.70 12.4
bench_composite 0.30 23.4

benchmark panel evidence:

complete panelscore (0–100)
swe-bench-verified-v123.4

Coding: provisional — score 5.6 ● low — a single signal; treat with caution

provisional: too few independent signals for a numbered position — the score is shown, but this model sorts after every ranked model.

signalweightinput (0–100)
aa_coding 0.30
aider_polyglot 0.30
swe_bench_verified 0.40 5.6

missing signals are dropped and the remaining weights renormalized — never imputed.

Agentic: #20 — score 5.6 ● low — a single signal; treat with caution

signalweightinput (0–100)
swe_bench_verified 0.60 5.6
swe_rebench 0.40

missing signals are dropped and the remaining weights renormalized — never imputed.

full methodology

# trend

OGM score · last 55 days
OGM score over 55 days
overall rank (up is better)
Overall rank over 55 days

methodology changed during this history window; score movement across that boundary is not model movement. See methodology v5.

# benchmarks

benchmarkscoresourcedate
AA Intelligence Index 3.6 Artificial Analysis
AA Math Index 29.3 Artificial Analysis
SWE-bench Verified 38.0 SWE-bench Jul 2025

# where to run

providerquantctx$/M in$/M out$/M cacheprice srctpsuptime
Mistral 125k $0.10$0.30 via modelsdev

sorted by blended price ((3·input + output) / 4 per 1M) · ✓ = the provider's own catalog confirms the offer · "via …" prices are what the aggregator routing the offer charges, not the provider's own list price

source aliases
aa
devstral-small
openrouter
mistralai/devstral-small