Devstral Small 2
- Input
- —
- Output
- —
- Speed
- ~99tok/sest.
- Context
- 256Ktokens
Long-tail import from Artificial Analysis (2026-09-24); AA variant: Devstral Small 2. Price = AA median host price.
Rankings
Coding
Verified#126 / 137rank by quality21.2±6.5quality / 1003/7 benchmarks measuredWriting & chat
Verified#134 / 162rank by quality20.8±9.5quality / 1001/5 benchmarks measuredVision
Verified#86 / 98rank by quality20.1±6.3quality / 1002/6 benchmarks measuredHard reasoning
Verified#175 / 218rank by quality18.6±3.3quality / 1004/4 benchmarks measuredAgents
Verified#136 / 187rank by quality17.8±4.3quality / 1004/7 benchmarks measured
Self-hostingEstimated
How we estimate →What it would cost to run these open weights yourself on rented GPUs. No API sells this model, so this is its price on the chart.
- Hardware
- 1× B200 180GB
- BF16 weights
- Throughput
- ~4,962 tok/s
- many requests batched
- Price
- $0.19–1.67 /M tok
- busy → light use
- On your own machine
- RTX 5090 32GB (int4, ~100 tok/s single-stream)
- single consumer GPU or Mac
Assumptions (5)
- BF16 weights (24B params) + 25% KV-cache headroom ≈ 60 GB
- 1× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
- ~4,962 output tok/s aggregate at batch 142 (bandwidth-bound)
- Blended 3:1 input:output; prefill ~22,148 tok/s
- Low = 75% utilization at the low GPU price; high = 20% at the high price
Benchmark scores
~ italic, dashed = no published score yet, estimated from related benchmarks.
Coding
SWE-bench VerifiedEstimated
~#17 of 21 (estimated)~64.9%
- #1Claude Opus 596.0%
- #2Claude Sonnet 585.2%
- #16Nemotron 3 Ultra70.7%
- ~#17Devstral Small 2Estimated64.9%
- #17Intern-S2-Preview (35B-A3B)64.0%
No published score — estimated from related benchmarks
SWE-bench Pro (Public)Estimated
~#23 of 24 (estimated)~44.6%
- #1Claude Fable 5.293.0%
- #2Claude Opus 5.589.9%
- #22Muse Glimmer51.2%
- ~#23Devstral Small 2Estimated44.6%
- #23Nex-N2.5-Mini43.8%
No published score — estimated from related benchmarks
DeepSWE v1.1Estimated
~#43 of 46 (estimated)~35.5%
- #1Gemini 4 Argon78.8%
- #2DeepSeek V4.1 Flash 101777.0%
- #42Nex-N2.5-Mini36.1%
- ~#43Devstral Small 2Estimated35.5%
- #43Hy328.0%
No published score — estimated from related benchmarks
Terminal-Bench 4.0Verified
#63 of 1010.0%
- #1Kimi K3.1 Pro67.2%
- #2Claude Sonnet 5.563.6%
- #63Devstral 20.0%
- #63Devstral Small 20.0%
- #63Gemma 4 31B0.0%
artificialanalysis.ai · 2026-09-24
LiveCodeBenchVerified
#24 of 2734.8%
- #1Qwen3.8-Omni-Flash92.6%
- #2Qwen3.8 Flash-Next91.9%
- #23Ministral 3 14B35.1%
- #24Devstral Small 234.8%
- #25Ministral 3 8B30.3%
artificialanalysis.ai · 2026-09-24
SciCodeVerified
#100 of 11832.4%
- #1Claude Fable 5.267.7%
- #2Claude Opus 5.566.9%
- #99Devstral 232.8%
- #100Devstral Small 232.4%
- #101Nemotron 3.5 Lightning32.1%
artificialanalysis.ai · 2026-09-24
AA Coding Agent IndexEstimated
~#25 of 26 (estimated)~27.4
- #1Claude Opus 5.561.7
- #2Claude Opus 559.7
- #24GPT-6 Luna31.1
- ~#25Devstral Small 2Estimated27.4
- #25Muse Spark Open26.3
No published score — estimated from related benchmarks
Writing & chat
IFBenchVerified
#113 of 13931.2%
- #1MiniMax-M382.9%
- #2Qwen3.8 2.4T A95B82.8%
- #112Jamba 1.7 Mini31.4%
- #113Devstral Small 231.2%
- #113Qwen3 Omni 30B A3B Instruct31.2%
artificialanalysis.ai · 2026-09-24
MMMLUEstimated
~last of 5 (estimated)below measured range (<81.3%)
No published score — estimated from related benchmarks
LMArena Text, Non-English (Elo)Estimated
~#48 of 57 (estimated)~1317 Elo
- #1Gemini 4 ArgonFew votes1512
- #2Claude Opus 5.51497
- #47INTELLECT-3Few votes1327
- ~#48Devstral Small 2Estimated1317
- #48Llama 4 Maverick1311
No published score — estimated from related benchmarks
LMArena Text (Elo)Estimated
~#49 of 57 (estimated)~1341 Elo
- #1Gemini 4 Argon1525
- #2Claude Opus 5.51507
- #48Llama 3.1 Nemotron Ultra 253B v1Few votes1348
- ~#49Devstral Small 2Estimated1341
- #49Olmo 3.1 32B Instruct1330
No published score — estimated from related benchmarks
EQ-Bench Creative Writing v3 (Elo)Estimated
~#24 of 26 (estimated)~1252 Elo
- #1GPT-6 Astra2173
- #2Claude Fable 5.12162
- #23Gemma 4 31B1368
- ~#24Devstral Small 2Estimated1252
- #24gpt-oss-120b961
No published score — estimated from related benchmarks
Vision
MMMU-ProVerified
#75 of 8744.6%
- #1Claude Opus 5.587.7%
- #2GPT-6 Astra86.9%
- #74Ministral 3 8B46.0%
- #75Devstral Small 244.6%
- #76Qwen3.5 2B42.7%
artificialanalysis.ai · 2026-09-24
CharXiv ReasoningEstimated
~last of 12 (estimated)below measured range (<74.5%)
- #1Gemini 3.8 Flash86.2%
- #2Kimi K384.8%
- #11Gemini 3.5 Flash-Lite74.5%
- ~#12Devstral Small 2Estimated<74.5%
No published score — estimated from related benchmarks
OmniDocBench v1.5Estimated
~last of 8 (estimated)below measured range (<75.8%)
No published score — estimated from related benchmarks
GDP.pdf (AA)Verified
#43 of 471.0%
- #1Muse Wonder Full32.6%
- #2GPT-6 Astra31.0%
- #43Command A+1.0%
- #43Devstral Small 21.0%
- #45Ministral 3 8B0.8%
artificialanalysis.ai · 2026-09-24
ChartographyEstimated
~last of 25 (estimated)below measured range (<9.0%)
No published score — estimated from related benchmarks
Hard reasoning
GPQA DiamondVerified
#149 of 19253.2%
- #1GPT-6 Astra96.1%
- #2Gemini 3.8 Flash95.3%
- #148Olmo 3.1 32B Instruct53.9%
- #149Devstral Small 253.2%
- #150Reka Flash 352.9%
artificialanalysis.ai · 2026-09-24
Humanity's Last ExamVerified
#196 of 1983.5%
- #1Claude Fable 5.274.3%
- #2Claude Opus 5.561.4%
- #192Llama 3.3 Instruct 70B3.6%
- #196Devstral Small 23.5%
- #197ERNIE 4.5 300B A47B3.3%
artificialanalysis.ai · 2026-09-24
Agents
τ²-bench (Telecom)Verified
#74 of 11923.4%
- #1Step 3.7 Flash98.5%
- #2DeepSeek-V4-Pro-0813-nvfp4-DSpark98.3%
- #73Mistral Large 324.6%
- #74Devstral Small 223.4%
- #74NVIDIA Nemotron Nano 9B V223.4%
artificialanalysis.ai · 2026-09-24
OSWorld-VerifiedEstimated
~#12 of 15 (estimated)~58.0%
- #1Kimi K384.8%
- #2Qwen3.8 27B84.3%
- #11Muse Glimmer65.9%
- ~#12Devstral Small 2Estimated58.0%
- #12Nemotron 3.5 Super VL NVFP452.3%
No published score — estimated from related benchmarks
OSWorld 2.0Estimated
~last of 5 (estimated)below measured range (<47.5%)
No published score — estimated from related benchmarks
BrowseCompEstimated
~#19 of 22 (estimated)~51.0%
- #1Atria Dawn Preview92.5%
- #2GPT-6 Astra91.5%
- #18Nemotron 3.5 Bolt 74B-A9B70.0%
- ~#19Devstral Small 2Estimated51.0%
- #19Mistral Medium 3.548.6%
No published score — estimated from related benchmarks
GDPval-AA (Elo)Verified
#67 of 91516 Elo
- #1Claude Opus 5.51866
- #2Muse Wonder Full1857
- #66Devstral 2531
- #67Devstral Small 2516
- #68Qwen3 Coder Next501
artificialanalysis.ai · 2026-09-24