Ministral 3 14B
- Input
- $0.20/M tok
- Output
- $0.20/M tok
- Blended
- $0.20/M tok · 3:1
- Speed
- 58tok/s
- Context
- 256Ktokens
- Cost per task
- ~$0.08/taskest.
Long-tail import from Artificial Analysis (2026-09-24); AA variant: Ministral 3 14B. Price = AA median host price.
Rankings
Coding
Verified#132 / 137rank by quality17.3±6.5quality / 1003/7 benchmarks measuredWriting & chat
Verified#133 / 162rank by quality21.2±9.5quality / 1001/5 benchmarks measuredVision
Verified#84 / 98rank by quality23.0±6.3quality / 1002/6 benchmarks measuredHard reasoning
Verified#176 / 218rank by quality18.5±3.3quality / 1004/4 benchmarks measuredAgents
Verified#144 / 187rank by quality15.9±4.3quality / 1004/7 benchmarks measured
Self-hostingEstimated
How we estimate →What it would cost to run these open weights yourself on rented GPUs, compared with the API price above.
- Hardware
- 1× B200 180GB
- BF16 weights
- Throughput
- ~7,655 tok/s
- many requests batched
- Price
- $0.12–1.04 /M tok
- busy → light use
- On your own machine
- RTX 5090 32GB (int4, ~180 tok/s single-stream)
- single consumer GPU or Mac
Assumptions (5)
- BF16 weights (13.95B params) + 25% KV-cache headroom ≈ 35 GB
- 1× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
- ~7,655 output tok/s aggregate at batch 219 (bandwidth-bound)
- Blended 3:1 input:output; prefill ~38,105 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
~#19 of 21 (estimated)~61.1%
- #1Claude Opus 596.0%
- #2Claude Sonnet 585.2%
- #18gpt-oss-120b62.4%
- ~#19Ministral 3 14BEstimated61.1%
- #19NVIDIA Nemotron-Labs Dory1.0 4B59.5%
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%
- ~#23Ministral 3 14BEstimated44.6%
- #23Nex-N2.5-Mini43.8%
No published score — estimated from related benchmarks
DeepSWE v1.1Estimated
~#43 of 46 (estimated)~31.5%
- #1Gemini 4 Argon78.8%
- #2DeepSeek V4.1 Flash 101777.0%
- #42Nex-N2.5-Mini36.1%
- ~#43Ministral 3 14BEstimated31.5%
- #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%
- #63MiniCPM5-2B0.0%
- #63Ministral 3 14B0.0%
- #63Ministral 3 3B0.0%
artificialanalysis.ai · 2026-09-24
LiveCodeBenchVerified
#23 of 2735.1%
- #1Qwen3.8-Omni-Flash92.6%
- #2Qwen3.8 Flash-Next91.9%
- #22Llama 4 Maverick39.7%
- #23Ministral 3 14B35.1%
- #24Devstral Small 234.8%
artificialanalysis.ai · 2026-09-24
SciCodeVerified
#110 of 11823.8%
- #1Claude Fable 5.267.7%
- #2Claude Opus 5.566.9%
- #109Ling 3.0 Tiny24.2%
- #110Ministral 3 14B23.8%
- #111K2 Horizon 3.7B22.0%
artificialanalysis.ai · 2026-09-24
AA Coding Agent IndexEstimated
~#25 of 26 (estimated)~27.3
- #1Claude Opus 5.561.7
- #2Claude Opus 559.7
- #24GPT-6 Luna31.1
- ~#25Ministral 3 14BEstimated27.3
- #25Muse Spark Open26.3
No published score — estimated from related benchmarks
Writing & chat
IFBenchVerified
#109 of 13932.0%
- #1MiniMax-M382.9%
- #2Qwen3.8 2.4T A95B82.8%
- #108Hermes 4 - Llama-3.1 405B32.7%
- #109Ministral 3 14B32.0%
- #110NVIDIA Nemotron Nano 12B v2 VL31.9%
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
- ~#48Ministral 3 14BEstimated1317
- #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
- ~#49Ministral 3 14BEstimated1341
- #49Olmo 3.1 32B Instruct1330
No published score — estimated from related benchmarks
EQ-Bench Creative Writing v3 (Elo)Estimated
~#24 of 26 (estimated)~1254 Elo
- #1GPT-6 Astra2173
- #2Claude Fable 5.12162
- #23Gemma 4 31B1368
- ~#24Ministral 3 14BEstimated1254
- #24gpt-oss-120b961
No published score — estimated from related benchmarks
Vision
MMMU-ProVerified
#73 of 8749.8%
- #1Claude Opus 5.587.7%
- #2GPT-6 Astra86.9%
- #72Gemma 4 E4B51.2%
- #73Ministral 3 14B49.8%
- #74Ministral 3 8B46.0%
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%
- ~#12Ministral 3 14BEstimated<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
#41 of 471.2%
- #1Muse Wonder Full32.6%
- #2GPT-6 Astra31.0%
- #40Mistral Medium 3.52.8%
- #41Ministral 3 14B1.2%
- #41Mistral Large 31.2%
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
#141 of 19257.2%
- #1GPT-6 Astra96.1%
- #2Gemini 3.8 Flash95.3%
- #140Phi-457.5%
- #141Ministral 3 14B57.2%
- #141NVIDIA Nemotron Nano 12B v2 VL57.2%
artificialanalysis.ai · 2026-09-24
Humanity's Last ExamVerified
#174 of 1984.6%
- #1Claude Fable 5.274.3%
- #2Claude Opus 5.561.4%
- #173Gemma 4 E2B4.7%
- #174Ministral 3 14B4.6%
- #174Nova Micro4.6%
artificialanalysis.ai · 2026-09-24
Agents
τ²-bench (Telecom)Verified
#65 of 11927.2%
- #1Step 3.7 Flash98.5%
- #2DeepSeek-V4-Pro-0813-nvfp4-DSpark98.3%
- #62Magistral Small 1.227.8%
- #65Ministral 3 14B27.2%
- #66INTELLECT-326.6%
artificialanalysis.ai · 2026-09-24
OSWorld-VerifiedEstimated
~#12 of 15 (estimated)~56.9%
- #1Kimi K384.8%
- #2Qwen3.8 27B84.3%
- #11Muse Glimmer65.9%
- ~#12Ministral 3 14BEstimated56.9%
- #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)~49.5%
- #1Atria Dawn Preview92.5%
- #2GPT-6 Astra91.5%
- #18Nemotron 3.5 Bolt 74B-A9B70.0%
- ~#19Ministral 3 14BEstimated49.5%
- #19Mistral Medium 3.548.6%
No published score — estimated from related benchmarks
GDPval-AA (Elo)Verified
#86 of 91234 Elo
- #1Claude Opus 5.51866
- #2Muse Wonder Full1857
- #85Nemotron Cascade 2 30B A3B252
- #86Ministral 3 14B234
- #87Nemotron 3 Nano Omni 30B A3B Reasoning216
artificialanalysis.ai · 2026-09-24