Skip to content

EthenEthenEthen

LoginPlatformTry Ethen

Mistral · Flagships Analysis

Devstral Small

Intelligence, Performance & Price Analysis

Back to catalogMore from Mistral

Canonical slug: devstral-small · Canonical model registry at build time

Intelligence

9 score

Speed

70.2 output tokens/sec

Latency

1.13s TTFT

Input Price

$0.10 / 1M tokens

Output Price

$0.30 / 1M tokens

Executive Assessment

Routing Verdict & Tradeoffs

Ethen Routing Verdict

Budget-friendly / task-specific model

Devstral Small (Jul '25) scores 9 (estimated) on the Artificial Analysis Intelligence Index, placing it well above average among other open weight non-reasoning models of similar size (median: 6). Devstral Small (Jul '25) generates output at 70.2 tokens per second (based on Mistral's API), which is below average compared to other open weight non-reasoning models of similar size (median: 102.7 t/s). Devstral Small (Jul '25) costs $0.10 per 1M input tokens (better than average, median: $0.15) and $0.30 per 1M output tokens (better than average, median: $0.32), based on Mistral's API.

Best for high-volume, simple, or domain-specific tasks where cost or speed matters more than deep reasoning.

Task Fit Assessment

WorkloadRatingNotes
Complex reasoning & agentic workflowssuboptimalIntelligence score 9 is better suited for straightforward tasks than multi-step reasoning.
High-volume chat & customer-facingoptimalOutput speed 70.2 tokens/sec is adequate for chat.
Latency-sensitive applicationsoptimalTTFT 1.13s — adequate latency for most interactive use cases.
Cost-sensitive pipelinesoptimalOutput pricing at $0.30 is very competitive for high-volume workloads.

Cost Pressure Analysis

Low

Pricing is competitive — input $0.10, output $0.30. Suitable for sustained production use.

Technical Specifications

Architecture and Limits

SpecificationValueValidation Authority
Model typeOpen weightsinferred
ReasoningNofaq
Input modalitiesDevstral Small (Jul '25) supports text input.faq
Output modalitiesDevstral Small (Jul '25) supports text output.faq
Context window260k tokensfaq
Open weights / sourceYes, Devstral Small (Jul '25) is open weights. The model weights are publicly available and can be downloaded for self-hosting.faq
ParametersDevstral Small (Jul '25) has 24 billion parameters.faq
LicenseDevstral Small (Jul '25) is released under the Apache 2.0 license. This license allows commercial use.faq
API availabilityYes, Devstral Small (Jul '25) is available via API through 1 provider.faq

Evidence charts

Profile visualizations

Charts are the same canonical evidence cards previously published for this slug, contained inside the D18D page grammar.

AA-Omniscience Index

AA-Omniscience Index (higher is better) measures knowledge reliability and hallucination. It rewards correct answers, penalizes hallucinations, and has no penalty for refusing to answer. Scores range from -100 to 100, where 0 means as many correct as incorrect answers, and negative scores mean more incorrect than correct. · Evaluation results measured independently by Artificial Analysis

Retrieved 2026-07-08

Chart source and provenance are listed in Methodology & sources below.

Retrieval date:
2026-07-08
Methodology:
Independent test run by Artificial Analysis on dedicated hardware.

-100-50050Claude Fable 5 (with fallback): 4040AClaude Fable 5 (with fal…Gemini 3.5 Flash: 2323GGemini 3.5 FlashGPT-5.5 (xhigh): 2020AIGPT-5.5 (xhigh)Grok 4.3 (high): 1818xGrok 4.3 (high)Kimi K2.6: 6.46.4KKimi K2.6Muse Spark: 4.14.1?Muse SparkGLM-5.2 (max): 44?GLM-5.2 (max)MiniMax-M3: 1.41.4?MiniMax-M3Nemotron 3 Ultra: -0.8-0.8NNemotron 3 UltraDeepSeek V4 Pro (max): -10-10DDeepSeek V4 Pro (max)Devstral Small: -51.1-51.1?Devstral Small

Artificial Analysis Intelligence Index by Open Weights / Proprietary

Artificial Analysis Intelligence Index v4.1 incorporates 9 evaluations: GDPval-AA v2, 𝜏³-Banking, Terminal-Bench v2.1, SciCode, Humanity's Last Exam, GPQA Diamond, CritPt, AA-Omniscience, AA-LCR · Evaluation results measured independently by Artificial Analysis

Retrieved 2026-07-08

Chart source and provenance are listed in Methodology & sources below.

Retrieval date:
2026-07-08
Methodology:
Independent test run by Artificial Analysis on dedicated hardware.

0204060Claude Fable 5 (with fallback): 6060AClaude Fable 5 (with fal…GPT-5.5 (xhigh): 5555AIGPT-5.5 (xhigh)GLM-5.2 (max): 5151?GLM-5.2 (max)Gemini 3.5 Flash: 5050GGemini 3.5 FlashMiniMax-M3: 4444?MiniMax-M3DeepSeek V4 Pro (max): 4444DDeepSeek V4 Pro (max)Kimi K2.6: 4444KKimi K2.6Muse Spark: 4343?Muse SparkNemotron 3 Ultra: 3838NNemotron 3 UltraGrok 4.3 (high): 3838xGrok 4.3 (high)Devstral Small 2: 1717?Devstral Small 2

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index v4.1 incorporates 9 evaluations: GDPval-AA v2, 𝜏³-Banking, Terminal-Bench v2.1, SciCode, Humanity's Last Exam, GPQA Diamond, CritPt, AA-Omniscience, AA-LCR · Evaluation results measured independently by Artificial Analysis

Retrieved 2026-07-08

Chart source and provenance are listed in Methodology & sources below.

Retrieval date:
2026-07-08
Methodology:
Independent test run by Artificial Analysis on dedicated hardware.

0204060Claude Fable 5 (with fallback): 6060AClaude Fable 5 (with fal…GPT-5.5 (xhigh): 5555AIGPT-5.5 (xhigh)GLM-5.2 (max): 5151?GLM-5.2 (max)Gemini 3.5 Flash: 5050GGemini 3.5 FlashMiniMax-M3: 4444?MiniMax-M3DeepSeek V4 Pro (max): 4444DDeepSeek V4 Pro (max)Kimi K2.6: 4444KKimi K2.6Muse Spark: 4343?Muse SparkNemotron 3 Ultra: 3838NNemotron 3 UltraGrok 4.3 (high): 3838xGrok 4.3 (high)Devstral Small 2: 1717?Devstral Small 2

Artificial Analysis Openness Index: Score

Openness Index assesses model openness on a 0 to 100 normalized scale (higher is more open) · Evaluation results measured independently by Artificial Analysis

Retrieved 2026-07-08

Chart source and provenance are listed in Methodology & sources below.

Retrieval date:
2026-07-08
Methodology:
Independent test run by Artificial Analysis on dedicated hardware.

050100Nemotron 3 Ultra: 8383NNemotron 3 UltraGranite 4.1 30B: 6161?Granite 4.1 30BDeepSeek V4 Pro (max): 5050DDeepSeek V4 Pro (max)Gemma 3 27B: 5050?Gemma 3 27BGemma 3 12B: 5050?Gemma 3 12BDevstral Small: 4444?Devstral SmallGLM-5.2 (max): 4444?GLM-5.2 (max)gpt-oss-120b (high): 3939AIgpt-oss-120b (high)Qwen3.6 27B: 3939QQwen3.6 27BMistral Small 3.2: 3939MiMistral Small 3.2

Intelligence

Artificial Analysis Intelligence Index · Higher is better · Evaluation results measured independently by Artificial Analysis

Retrieved 2026-07-08

Chart source and provenance are listed in Methodology & sources below.

Retrieval date:
2026-07-08
Methodology:
Independent test run by Artificial Analysis on dedicated hardware.

0204060Claude Fable 5 (with fallback): 6060AClaude Fable 5 (with fal…GPT-5.5 (xhigh): 5555AIGPT-5.5 (xhigh)GLM-5.2 (max): 5151?GLM-5.2 (max)Gemini 3.5 Flash: 5050GGemini 3.5 FlashMiniMax-M3: 4444?MiniMax-M3DeepSeek V4 Pro (max): 4444DDeepSeek V4 Pro (max)Kimi K2.6: 4444KKimi K2.6Muse Spark: 4343?Muse SparkNemotron 3 Ultra: 3838NNemotron 3 UltraGrok 4.3 (high): 3838xGrok 4.3 (high)Devstral Small: 9.39.3?Devstral Small

Output Speed

Output tokens per second · Higher is better · Evaluation results measured independently by Artificial Analysis

Retrieved 2026-07-08

Chart source and provenance are listed in Methodology & sources below.

Retrieval date:
2026-07-08
Methodology:
Independent test run by Artificial Analysis on dedicated hardware.

088175263350gpt-oss-120b (high): 316316AIgpt-oss-120b (high)Nemotron 3 Ultra: 248248NNemotron 3 UltraGLM-5.2 (max): 216216?GLM-5.2 (max)Gemini 3.5 Flash: 192192GGemini 3.5 FlashMistral Small 3.1: 171171MiMistral Small 3.1Grok 4.3 (high): 164164xGrok 4.3 (high)Mistral Small 3.2: 147147MiMistral Small 3.2Llama 3.1 8B: 144144MLlama 3.1 8BMiniMax-M3: 9595?MiniMax-M3Ministral 3 8B: 9393?Ministral 3 8BDevstral Small: 7070?Devstral Small

Speed

Output tokens per second · Higher is better · Evaluation results measured independently by Artificial Analysis

Retrieved 2026-07-08

Chart source and provenance are listed in Methodology & sources below.

Retrieval date:
2026-07-08
Methodology:
Independent test run by Artificial Analysis on dedicated hardware.

088175263350gpt-oss-120b (high): 316316AIgpt-oss-120b (high)Nemotron 3 Ultra: 248248NNemotron 3 UltraGLM-5.2 (max): 216216?GLM-5.2 (max)Gemini 3.5 Flash: 192192GGemini 3.5 FlashGrok 4.3 (high): 164164xGrok 4.3 (high)MiniMax-M3: 9595?MiniMax-M3GPT-5.5 (xhigh): 8888AIGPT-5.5 (xhigh)Kimi K2.6: 7676KKimi K2.6DeepSeek V4 Pro (max): 7272DDeepSeek V4 Pro (max)Claude Fable 5 (with fallback): 7171AClaude Fable 5 (with fal…Devstral Small: 7070?Devstral Small

End-to-End Response Time

Seconds to output 500 tokens, including reasoning model 'thinking' time · Lower is better · Evaluation results measured independently by Artificial Analysis

Retrieved 2026-07-08

Chart source and provenance are listed in Methodology & sources below.

Retrieval date:
2026-07-08
Methodology:
Independent test run by Artificial Analysis on dedicated hardware.

0s2.5s5s7.5s10sQwen3.6 27B: 8.4s8.4sQQwen3.6 27BDevstral Small 2: 7.4s7.4s?Devstral Small 2Devstral Small: 7.1s7.1s?Devstral SmallClaude Fable 5 (with fallback): 7.1s7.1sAClaude Fable 5 (with fal…DeepSeek V4 Pro (max): 6.9s6.9sDDeepSeek V4 Pro (max)Kimi K2.6: 6.6s6.6sKKimi K2.6Ministral 3 14B: 6.4s6.4s?Ministral 3 14BGPT-5.5 (xhigh): 5.7s5.7sAIGPT-5.5 (xhigh)Ministral 3 8B: 5.4s5.4s?Ministral 3 8BMiniMax-M3: 5.2s5.2s?MiniMax-M3

Latency: Time To First Answer Token

Seconds to first answer token received · Accounts for reasoning model 'thinking' time · Evaluation results measured independently by Artificial Analysis

Retrieved 2026-07-08

Chart source and provenance are listed in Methodology & sources below.

Retrieval date:
2026-07-08
Methodology:
Independent test run by Artificial Analysis on dedicated hardware.

0s25s50s75s100sDeepSeek V4 Pro (max): 61s61sDDeepSeek V4 Pro (max)Kimi K2.6: 58s58sKKimi K2.6MiniMax-M3: 21s21s?MiniMax-M3GLM-5.2 (max): 9.3s9.3s?GLM-5.2 (max)Nemotron 3 Ultra: 9.2s9.2sNNemotron 3 Ultragpt-oss-120b (high): 6.3s6.3sAIgpt-oss-120b (high)Mistral Small 3.2: 0s0sMiMistral Small 3.2Mistral Small 3.1: 0s0sMiMistral Small 3.1Ministral 3 8B: 0s0s?Ministral 3 8BMinistral 3 14B: 0s0s?Ministral 3 14BDevstral Small: 0s0s?Devstral Small

Context Window

Context window: tokens limit · Higher is better · Evaluation results measured independently by Artificial Analysis

Retrieved 2026-07-08

Chart source and provenance are listed in Methodology & sources below.

Retrieval date:
2026-07-08
Methodology:
Independent test run by Artificial Analysis on dedicated hardware.

0250k500k750k1MGemini 3.5 Flash: 1M1MGGemini 3.5 FlashClaude Fable 5 (with fallback): 1M1MAClaude Fable 5 (with fal…GLM-5.2 (max): 1M1M?GLM-5.2 (max)DeepSeek V4 Pro (max): 1M1MDDeepSeek V4 Pro (max)Grok 4.3 (high): 1M1MxGrok 4.3 (high)MiniMax-M3: 1M1M?MiniMax-M3GPT-5.5 (xhigh): 922k922kAIGPT-5.5 (xhigh)Nemotron 3 Ultra: 262k262kNNemotron 3 UltraMuse Spark: 262k262k?Muse SparkQwen3.6 27B: 262k262kQQwen3.6 27BDevstral Small: 256k256k?Devstral Small

Cost per Intelligence Index Task

Weighted average cost (USD) per Artificial Analysis Intelligence Index task, segmented by token type. Lower is better · Evaluation results measured independently by Artificial Analysis

Retrieved 2026-07-08

Chart source and provenance are listed in Methodology & sources below.

Retrieval date:
2026-07-08
Methodology:
Independent test run by Artificial Analysis on dedicated hardware.

00.51Claude Fable 5 (with fallback): 0.40.4AClaude Fable 5 (with fal…GPT-5.5 (xhigh): 0.10.1AIGPT-5.5 (xhigh)Gemini 3.5 Flash: 0.10.1GGemini 3.5 FlashQwen3.6 27B: 00QQwen3.6 27BGLM-5.2 (max): 00?GLM-5.2 (max)Kimi K2.6: 00KKimi K2.6Nemotron 3 Ultra: 00NNemotron 3 UltraMiniMax-M3: 00?MiniMax-M3Grok 4.3 (high): 00xGrok 4.3 (high)DeepSeek V4 Pro (max): 00DDeepSeek V4 Pro (max)

Cost per Task

Weighted average cost (USD) per Intelligence Index task · Lower is better · Evaluation results measured independently by Artificial Analysis

Retrieved 2026-07-08

Chart source and provenance are listed in Methodology & sources below.

Retrieval date:
2026-07-08
Methodology:
Independent test run by Artificial Analysis on dedicated hardware.

$0.00$1.00$2.00$3.00$4.00$5.00Claude Fable 5 (with fallback): $2.75$2.75AClaude Fable 5 (with fal…GPT-5.5 (xhigh): $0.86$0.86AIGPT-5.5 (xhigh)Gemini 3.5 Flash: $0.59$0.59GGemini 3.5 FlashGLM-5.2 (max): $0.37$0.37?GLM-5.2 (max)Kimi K2.6: $0.35$0.35KKimi K2.6Nemotron 3 Ultra: $0.24$0.24NNemotron 3 UltraGrok 4.3 (high): $0.14$0.14xGrok 4.3 (high)MiniMax-M3: $0.12$0.12?MiniMax-M3gpt-oss-120b (high): $0.06$0.06AIgpt-oss-120b (high)DeepSeek V4 Pro (max): $0.04$0.04DDeepSeek V4 Pro (max)

Cost to Run Artificial Analysis Intelligence Index

Cost (USD) to run all evaluations in the Artificial Analysis Intelligence Index · Evaluation results measured independently by Artificial Analysis

Retrieved 2026-07-08

Chart source and provenance are listed in Methodology & sources below.

Retrieval date:
2026-07-08
Methodology:
Independent test run by Artificial Analysis on dedicated hardware.

0200400600Claude Fable 5 (with fallback): 508508AClaude Fable 5 (with fal…GPT-5.5 (xhigh): 132132AIGPT-5.5 (xhigh)Gemini 3.5 Flash: 114114GGemini 3.5 FlashQwen3.6 27B: 8888QQwen3.6 27BNemotron 3 Ultra: 3434NNemotron 3 UltraGLM-5.2 (max): 3333?GLM-5.2 (max)Kimi K2.6: 2525KKimi K2.6MiniMax-M3: 1515?MiniMax-M3Grok 4.3 (high): 1313xGrok 4.3 (high)DeepSeek V4 Pro (max): 7.27.2DDeepSeek V4 Pro (max)

Pricing: Cache Hit, Input, and Output

Price (USD per M Tokens) · Evaluation results measured independently by Artificial Analysis

Retrieved 2026-07-08

Chart source and provenance are listed in Methodology & sources below.

Retrieval date:
2026-07-08
Methodology:
Independent test run by Artificial Analysis on dedicated hardware.

0250k500k750k1MClaude Fable 5 (with fallback): 6161AClaude Fable 5 (with fal…GPT-5.5 (xhigh): 3636AIGPT-5.5 (xhigh)Gemini 3.5 Flash: 1111GGemini 3.5 FlashGLM-5.2 (max): 6.16.1?GLM-5.2 (max)Kimi K2.6: 5.15.1KKimi K2.6Qwen3.6 27B: 4.24.2QQwen3.6 27BGrok 4.3 (high): 44xGrok 4.3 (high)Nemotron 3 Ultra: 3.63.6NNemotron 3 UltraMiniMax-M3: 1.61.6?MiniMax-M3DeepSeek V4 Pro (max): 1.31.3DDeepSeek V4 Pro (max)Devstral Small: 0.40.4?Devstral Smallcache Hitinputoutput

Time per Intelligence Index Task

Weighted average decode time (minutes) per task; excludes TTFT and overhead time · Lower is better · Evaluation results measured independently by Artificial Analysis

Retrieved 2026-07-08

Chart source and provenance are listed in Methodology & sources below.

Retrieval date:
2026-07-08
Methodology:
Independent test run by Artificial Analysis on dedicated hardware.

0510Kimi K2.6: 8.28.2KKimi K2.6DeepSeek V4 Pro (max): 6.76.7DDeepSeek V4 Pro (max)Claude Fable 5 (with fallback): 4.84.8AClaude Fable 5 (with fal…MiniMax-M3: 44?MiniMax-M3Qwen3.6 27B: 33QQwen3.6 27BGPT-5.5 (xhigh): 33AIGPT-5.5 (xhigh)GLM-5.2 (max): 2.82.8?GLM-5.2 (max)Ministral 3 8B: 2.52.5?Ministral 3 8BMinistral 3 14B: 2.22.2?Ministral 3 14BGemini 3.5 Flash: 2.22.2GGemini 3.5 FlashDevstral Small 2: 1.41.4?Devstral Small 2

Model Size: Total and Active Parameters

Comparison between total model parameters and parameters active during inference · Evaluation results measured independently by Artificial Analysis

Retrieved 2026-07-08

Chart source and provenance are listed in Methodology & sources below.

Retrieval date:
2026-07-08
Methodology:
Independent test run by Artificial Analysis on dedicated hardware.

0250k500k750k1MDeepSeek V4 Pro (max): 1.6k1.6kDDeepSeek V4 Pro (max)Kimi K2.6: 1k1kKKimi K2.6GLM-5.2 (max): 753753?GLM-5.2 (max)Nemotron 3 Ultra: 550550NNemotron 3 UltraMiniMax-M3: 428428?MiniMax-M3gpt-oss-120b (high): 117117AIgpt-oss-120b (high)EXAONE 4.5 33B: 3434?EXAONE 4.5 33BGranite 4.1 30B: 3030?Granite 4.1 30BQwen3.6 27B: 2828QQwen3.6 27BGemma 3 27B: 2727?Gemma 3 27BDevstral Small: 2424?Devstral Smallpassive Paramsactive Params

Methodology

Methodology & Provenance

This page is rendered from the normalized profile and page JSON for Devstral Small.

Benchmark values are preserved as normalized; only layout, disclosure ordering, and typography are adjusted for readability.

Frequently Asked Questions

Model FAQs & Technical Disclosures

Devstral Small (Jul '25) was released on July 10, 2025.

Related

Related

Provider

Mistral

Other Mistral models in the catalog.

Open

Directory

Leaderboards

Certified metric ranks, when published.

Open