Skip to content

Mistral · Model profile

Ministral 3 3B

Intelligence, Performance & Price Analysis

Source: Canonical slug: ministral-3-3b · Canonical model registry at build time

Routing verdict

Budget-friendly / task-specific model

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

Ministral 3 3B summary metrics

Intelligence
7 scoreMinistral 3 3B scores 7 on the Artificial Analysis Intelligence Index, placing it well above average among other open weight non-reasoning models of similar size (median: 3).
Speed
178.1 output tokens/secMinistral 3 3B generates output at 178.1 tokens per second (based on Mistral's API), which is well above average compared to other open weight non-reasoning models of similar size (median: 85.1 t/s).
Latency
0.68s TTFTMinistral 3 3B has a time to first token (TTFT) of 0.68s (based on Mistral's API), which is very competitive compared to other open weight non-reasoning models of similar size (median: 0.84s).
Input Price
$0.10 / 1M tokensMinistral 3 3B costs $0.10 per 1M input tokens (very competitive, median: $0.11) and $0.10 per 1M output tokens (very competitive, median: $0.28), based on Mistral's API.
Output Price
$0.10 / 1M tokensMinistral 3 3B costs $0.10 per 1M input tokens (very competitive, median: $0.11) and $0.10 per 1M output tokens (very competitive, median: $0.28), based on Mistral's API.
Verbosity
27M Output tokens from Intelligence Index 2 out of 4 units for Verbosity . Compa

Assessment

Routing verdict and task fit

Ministral 3 3B scores 7 on the Artificial Analysis Intelligence Index, placing it well above average among other open weight non-reasoning models of similar size (median: 3). Ministral 3 3B generates output at 178.1 tokens per second (based on Mistral's API), which is well above average compared to other open weight non-reasoning models of similar size (median: 85.1 t/s). Ministral 3 3B costs $0.10 per 1M input tokens (very competitive, median: $0.11) and $0.10 per 1M output tokens (very competitive, median: $0.28), based on Mistral's API.
Task fit assessment
WorkloadRatingNotes
Complex reasoning & agentic workflowsSuboptimalIntelligence score 7 is better suited for straightforward tasks than multi-step reasoning.
High-volume chat & customer-facingOptimalOutput speed 178.1 tokens/sec is adequate for chat.
Latency-sensitive applicationsOptimalTTFT 0.68s — among the lowest latencies, suitable for interactive latency-critical use cases.
Cost-sensitive pipelinesOptimalOutput pricing at $0.10 is very competitive for high-volume workloads.

Evidence

Evidence charts

17 published comparison charts for this profile. Ministral 3 3B is marked with a diamond and bold label; each chart has a table view and lists its source and retrieval date.

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

16 models in the published order. Ministral 3 3B: -64.1, row 13 of 16 in the published order.
  1. Claude Fable 5 (with fallback)40.1
  2. Gemini 3.5 Flash22.7
  3. GPT-5.5 (xhigh)20.1
  4. Grok 4.3 (high)18.3
  5. Kimi K2.66.4
  6. Muse Spark4.1
  7. GLM-5.2 (max)4
  8. MiniMax-M31.4
  9. Nemotron 3 Ultra-0.833
  10. DeepSeek V4 Pro (max)-10
  11. gpt-oss-120b (high)-50
  12. Phi-4 Mini-61
  13. Ministral 3 3B (this model)-64.1
  14. Granite 4.1 3B-77.4
  15. Qwen3.5 2B-82.8
  16. MiniCPM-V 4.6 1.3B-84.7

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Chart interpretation
Each bar is one model measured the same way, so this model can be read against the others. The chart lists its source and retrieval date.

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

16 models in the published order. Ministral 3 3B: 6.8, row 12 of 16 in the published order.
  1. Claude Fable 5 (with fallback)59.9
  2. GPT-5.5 (xhigh)54.8
  3. GLM-5.2 (max)51.1
  4. Gemini 3.5 Flash50.2
  5. MiniMax-M344.4
  6. DeepSeek V4 Pro (max)44.3
  7. Kimi K2.644.2
  8. Muse Spark43.1
  9. Nemotron 3 Ultra37.8
  10. Grok 4.3 (high)37.6
  11. gpt-oss-120b (high)23.8
  12. Ministral 3 3B (this model)6.8
  13. Phi-4 Mini6
  14. Qwen3.5 2B5.6
  15. Granite 4.1 3B4.7
  16. MiniCPM-V 4.6 1.3B4.2

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Benchmark interpretation
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.

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

16 models in the published order. Ministral 3 3B: 6.8, row 12 of 16 in the published order.
  1. Claude Fable 5 (with fallback)59.9
  2. GPT-5.5 (xhigh)54.8
  3. GLM-5.2 (max)51.1
  4. Gemini 3.5 Flash50.2
  5. MiniMax-M344.4
  6. DeepSeek V4 Pro (max)44.3
  7. Kimi K2.644.2
  8. Muse Spark43.1
  9. Nemotron 3 Ultra37.8
  10. Grok 4.3 (high)37.6
  11. gpt-oss-120b (high)23.8
  12. Ministral 3 3B (this model)6.8
  13. Phi-4 Mini6
  14. Qwen3.5 2B5.6
  15. Granite 4.1 3B4.7
  16. MiniCPM-V 4.6 1.3B4.2

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Benchmark interpretation
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.

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

10 models in the published order. Ministral 3 3B is not marked in this published set of 10 models.
  1. Nemotron 3 Ultra83.3
  2. Granite 4.1 3B61.1
  3. DeepSeek V4 Pro (max)50
  4. Phi-4 Mini50
  5. GLM-5.2 (max)44.4
  6. MiniCPM-V 4.6 1.3B44.4
  7. gpt-oss-120b (high)38.9
  8. Qwen3.5 2B38.9
  9. Kimi K2.633.3
  10. MiniMax-M333.3

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Chart interpretation
Each bar is one model measured the same way, so this model can be read against the others. The chart lists its source and retrieval date.

Intelligence

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

12 models in the published order. Ministral 3 3B: 6.8, row 12 of 12 in the published order.
  1. Claude Fable 5 (with fallback)59.9
  2. GPT-5.5 (xhigh)54.8
  3. GLM-5.2 (max)51.1
  4. Gemini 3.5 Flash50.2
  5. MiniMax-M344.4
  6. DeepSeek V4 Pro (max)44.3
  7. Kimi K2.644.2
  8. Muse Spark43.1
  9. Nemotron 3 Ultra37.8
  10. Grok 4.3 (high)37.6
  11. gpt-oss-120b (high)23.8
  12. Ministral 3 3B (this model)6.8

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Benchmark interpretation
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.

Output Speed

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

12 models in the published order. Ministral 3 3B: 178 t/s, row 5 of 12 in the published order.
  1. gpt-oss-120b (high)314 t/s
  2. Nemotron 3 Ultra250 t/s
  3. GLM-5.2 (max)218 t/s
  4. Gemini 3.5 Flash192 t/s
  5. Ministral 3 3B (this model)178 t/s
  6. Grok 4.3 (high)164 t/s
  7. MiniMax-M399 t/s
  8. GPT-5.5 (xhigh)88 t/s
  9. Kimi K2.676 t/s
  10. DeepSeek V4 Pro (max)72 t/s
  11. Claude Fable 5 (with fallback)71 t/s
  12. Phi-4 Mini44 t/s

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Output speed context
Speed should be read together with latency and cost; fast output does not always mean better user-perceived performance.

Speed

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

11 models in the published order. Ministral 3 3B: 178 t/s, row 5 of 11 in the published order.
  1. gpt-oss-120b (high)314 t/s
  2. Nemotron 3 Ultra250 t/s
  3. GLM-5.2 (max)218 t/s
  4. Gemini 3.5 Flash192 t/s
  5. Ministral 3 3B (this model)178 t/s
  6. Grok 4.3 (high)164 t/s
  7. MiniMax-M399 t/s
  8. GPT-5.5 (xhigh)88 t/s
  9. Kimi K2.676 t/s
  10. DeepSeek V4 Pro (max)72 t/s
  11. Claude Fable 5 (with fallback)71 t/s

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Output speed context
Speed should be read together with latency and cost; fast output does not always mean better user-perceived performance.

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

12 models in the published order. Ministral 3 3B: 2.81 s, row 1 of 12 in the published order.
  1. Ministral 3 3B (this model)2.81 s
  2. gpt-oss-120b (high)1.59 s
  3. Nemotron 3 Ultra2 s
  4. Phi-4 Mini11.44 s
  5. GLM-5.2 (max)2.3 s
  6. Grok 4.3 (high)3.05 s
  7. MiniMax-M35.05 s
  8. Gemini 3.5 Flash2.6 s
  9. Kimi K2.66.56 s
  10. DeepSeek V4 Pro (max)6.97 s
  11. GPT-5.5 (xhigh)5.68 s
  12. Claude Fable 5 (with fallback)7.07 s

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Chart interpretation
Each bar is one model measured the same way, so this model can be read against the others. The chart lists its source and retrieval date.

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

12 models in the published order. Ministral 3 3B: 0 s, row 1 of 12 in the published order.
  1. Ministral 3 3B (this model)0 s
  2. Phi-4 Mini0 s
  3. gpt-oss-120b (high)6.37 s
  4. Nemotron 3 Ultra9.09 s
  5. GLM-5.2 (max)9.19 s
  6. Grok 4.3 (high)0 s
  7. MiniMax-M320.19 s
  8. Gemini 3.5 Flash0 s
  9. Kimi K2.658.43 s
  10. DeepSeek V4 Pro (max)60.99 s
  11. GPT-5.5 (xhigh)0 s
  12. Claude Fable 5 (with fallback)0 s

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Latency considerations
Latency includes thinking / reasoning time when applicable. Lower is better for interactive use cases.

Context Window

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

16 models in the published order. Ministral 3 3B: 256k, row 12 of 16 in the published order.
  1. Gemini 3.5 Flash1M
  2. Claude Fable 5 (with fallback)1M
  3. GLM-5.2 (max)1M
  4. DeepSeek V4 Pro (max)1M
  5. Grok 4.3 (high)1M
  6. MiniMax-M31M
  7. GPT-5.5 (xhigh)922k
  8. Nemotron 3 Ultra262.1k
  9. Muse Spark262.1k
  10. Qwen3.5 2B262.1k
  11. MiniCPM-V 4.6 1.3B262k
  12. Ministral 3 3B (this model)256k
  13. Kimi K2.6256k
  14. gpt-oss-120b (high)131.1k
  15. Granite 4.1 3B131.1k
  16. Phi-4 Mini128k

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Context window sizing
Larger context windows enable complex multi-turn tasks but may increase latency and cost for long inputs.

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

11 models in the published order. Ministral 3 3B: 0.00204, row 4 of 11 in the published order.
  1. DeepSeek V4 Pro (max)0.0063
  2. gpt-oss-120b (high)0.00454
  3. MiniMax-M30.0139
  4. Ministral 3 3B (this model)0.00204
  5. Grok 4.3 (high)0.0101
  6. Nemotron 3 Ultra0.0209
  7. Kimi K2.60.0216
  8. GLM-5.2 (max)0.0268
  9. Gemini 3.5 Flash0.0884
  10. GPT-5.5 (xhigh)0.107
  11. Claude Fable 5 (with fallback)0.367

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Benchmark interpretation
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.
Cost pressure
Ethen Gateway should route away from this model when the task is simple, repetitive, or likely to generate many output tokens.

Cost per Task

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

11 models in the published order. Ministral 3 3B: $0.126, row 4 of 11 in the published order.
  1. DeepSeek V4 Pro (max)$0.045
  2. gpt-oss-120b (high)$0.061
  3. MiniMax-M3$0.125
  4. Ministral 3 3B (this model)$0.126
  5. Grok 4.3 (high)$0.139
  6. Nemotron 3 Ultra$0.244
  7. Kimi K2.6$0.346
  8. GLM-5.2 (max)$0.370
  9. Gemini 3.5 Flash$0.586
  10. GPT-5.5 (xhigh)$0.863
  11. Claude Fable 5 (with fallback)$2.75

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Cost pressure
Ethen Gateway should route away from this model when the task is simple, repetitive, or likely to generate many output tokens.

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

11 models in the published order. Ministral 3 3B: 2.7, row 1 of 11 in the published order.
  1. Ministral 3 3B (this model)2.7
  2. gpt-oss-120b (high)3.5
  3. DeepSeek V4 Pro (max)7.2
  4. MiniMax-M314.5
  5. Grok 4.3 (high)12.5
  6. Nemotron 3 Ultra33.5
  7. GLM-5.2 (max)32.9
  8. Kimi K2.625
  9. Gemini 3.5 Flash114
  10. GPT-5.5 (xhigh)131.8
  11. Claude Fable 5 (with fallback)507.6

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Benchmark interpretation
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.
Cost pressure
Ethen Gateway should route away from this model when the task is simple, repetitive, or likely to generate many output tokens.

Pricing: Cache Hit, Input, and Output

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

11 models; segments: Input, Output, Cache Hit. Ministral 3 3B: $0.200, row 1 of 11 in the published order.
  • Input
  • Output
  • Cache Hit
  1. Ministral 3 3B (this model)$0.200
  2. gpt-oss-120b (high)$0.900
  3. DeepSeek V4 Pro (max)$1.31
  4. MiniMax-M3$1.56
  5. Nemotron 3 Ultra$3.60
  6. Grok 4.3 (high)$3.95
  7. Kimi K2.6$5.11
  8. GLM-5.2 (max)$6.06
  9. Gemini 3.5 Flash$10.65
  10. GPT-5.5 (xhigh)$35.50
  11. Claude Fable 5 (with fallback)$61.00

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Chart interpretation
Each bar is one model measured the same way, so this model can be read against the others. The chart lists its source and retrieval date.

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

12 models in the published order. Ministral 3 3B: 1.8, row 3 of 12 in the published order.
  1. Nemotron 3 Ultra1.4
  2. Grok 4.3 (high)1.6
  3. Ministral 3 3B (this model)1.8
  4. gpt-oss-120b (high)1.9
  5. Gemini 3.5 Flash2.2
  6. GLM-5.2 (max)2.8
  7. GPT-5.5 (xhigh)3
  8. MiniMax-M33.9
  9. Claude Fable 5 (with fallback)4.8
  10. DeepSeek V4 Pro (max)6.8
  11. Kimi K2.68.2
  12. Phi-4 Mini10.7

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Benchmark interpretation
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.

Model Size: Total and Active Parameters

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

11 models; segments: Passive Params, Active Params. Ministral 3 3B: 3B, row 8 of 11 in the published order.
  • Passive Params
  • Active Params
  1. DeepSeek V4 Pro (max)1,600B
  2. Kimi K2.61,000B
  3. GLM-5.2 (max)753B
  4. Nemotron 3 Ultra550B
  5. MiniMax-M3428B
  6. gpt-oss-120b (high)117B
  7. Phi-4 Mini3.8B
  8. Ministral 3 3B (this model)3B
  9. Granite 4.1 3B3B
  10. Qwen3.5 2B2.3B
  11. MiniCPM-V 4.6 1.3B1.3B

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Architecture scale
Larger models generally perform better but cost more to serve. Active parameter count matters for inference efficiency.

Output Tokens per Intelligence Index Task

Weighted average number of output tokens used to run one task in the Artificial Analysis Intelligence Index · Evaluation results measured independently by Artificial Analysis

16 models in the published order. Ministral 3 3B: 20,441.6, row 6 of 16 in the published order.
  1. Granite 4.1 3B4,690.8
  2. Muse Spark2,857.2
  3. MiniCPM-V 4.6 1.3B13,811.5
  4. Grok 4.3 (high)4,032
  5. GPT-5.5 (xhigh)3,572.5
  6. Ministral 3 3B (this model)20,441.6
  7. Nemotron 3 Ultra7,796.7
  8. MiniMax-M311,623.4
  9. Gemini 3.5 Flash9,819.4
  10. Phi-4 Mini29,461.8
  11. Claude Fable 5 (with fallback)7,695.8
  12. gpt-oss-120b (high)7,571.4
  13. DeepSeek V4 Pro (max)7,243.2
  14. Kimi K2.65,397.6
  15. GLM-5.2 (max)6,089.1
  16. Qwen3.5 2B45,706.2

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Benchmark interpretation
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.

Specifications

Architecture and limits

Ministral 3 3B technical specifications
SpecificationValueSource
Model typeOpen weightsfaq
ReasoningNofaq
Input modalitiesMinistral 3 3B supports text and image input.faq
Output modalitiesMinistral 3 3B supports text output.faq
Context window260k tokensfaq
Open weights / sourceYes, Ministral 3 3B is open weights. The model weights are publicly available and can be downloaded for self-hosting.faq
ParametersMinistral 3 3B has 3 billion parameters.faq
LicenseMinistral 3 3B is released under the Apache 2.0 license. This license allows commercial use.faq
API availabilityYes, Ministral 3 3B is available via API through 2 providers.faq

Method

Methodology & Provenance

This page is rendered from the normalized profile and page JSON for Ministral 3 3B.

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

Questions

Model FAQ and technical disclosures

Ministral 3 3B was released on December 2, 2025.

Ministral 3 3B was created by Mistral.

Ministral 3 3B scores 7 on the Artificial Analysis Intelligence Index, placing it well above average among other open weight non-reasoning models of similar size (median: 3).

Ministral 3 3B generates output at 178.1 tokens per second (based on Mistral's API), which is well above average compared to other open weight non-reasoning models of similar size (median: 85.1 t/s).

Ministral 3 3B has a time to first token (TTFT) of 0.68s (based on Mistral's API), which is very competitive compared to other open weight non-reasoning models of similar size (median: 0.84s).

Ministral 3 3B costs $0.10 per 1M input tokens (very competitive, median: $0.11) and $0.10 per 1M output tokens (very competitive, median: $0.28), based on Mistral's API.

Ministral 3 3B costs $0.10 per 1M input tokens and $0.10 per 1M output tokens (based on Mistral's API). For a blended rate (7:2:1 cache hit/input/output ratio), this is $0.10 per 1M tokens. Pricing may vary by provider.

When evaluated on the Intelligence Index, Ministral 3 3B generated 27M output tokens, which is better than average compared to other open weight non-reasoning models of similar size (median: 27M).

No, Ministral 3 3B is not a reasoning model. It provides direct responses without extended chain-of-thought reasoning.

Ministral 3 3B supports text and image input.

Ministral 3 3B supports text output.

Yes, Ministral 3 3B supports image input and can analyze, describe, and answer questions about images.

Yes, Ministral 3 3B is multimodal. It can process text and image input and generate text output.

Ministral 3 3B has a context window of 260k tokens. This determines how much text and conversation history the model can process in a single request.

Yes, Ministral 3 3B is open weights. The model weights are publicly available and can be downloaded for self-hosting.

Ministral 3 3B has 3 billion parameters.

Ministral 3 3B is released under the Apache 2.0 license. This license allows commercial use.

Ministral 3 3B achieves a score of 7 on the Artificial Analysis Intelligence Index. This composite benchmark evaluates models across reasoning, knowledge, mathematics, and coding.

Yes, Ministral 3 3B is available via API through 2 providers.

Ministral 3 3B is available through 2 API providers.

Models

  • 01MistralOther Mistral models in the catalog.
  • 02LeaderboardsCertified metric ranks, when published.