Mistral Small 4
- Input
- $0.15/M tok
- Output
- $0.60/M tok
- Blended
- $0.262/M tok · 3:1
- Speed
- 170tok/s
- Context
- 256Ktokens
- Cost per task
- ~$0.25/taskest.
Open weights (Apache 2.0).
Rankings
Coding
Mixed#104 / 137rank by quality31.7±6.6quality / 1003/7 benchmarks measuredWriting & chat
Verified#96 / 162rank by quality35.9±9.5quality / 1001/5 benchmarks measuredVision
Verified#79 / 98rank by quality27.7±5.2quality / 1003/6 benchmarks measuredHard reasoning
Mixed#118 / 218rank by quality39.2±3.6quality / 1004/4 benchmarks measuredAgents
Mixed#166 / 187rank by quality13.2±3.4quality / 1005/7 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
- 2× H200 141GB
- FP8 weights
- Throughput
- ~8,963 tok/s
- many requests batched
- Price
- $0.088–0.73 /M tok
- busy → light use
- On your own machine
- Mac Studio M5 Max 128GB (int4, ~130 tok/s single-stream)
- single consumer GPU or Mac
Assumptions (5)
- FP8 weights (119B params, 6.5B active per token) + 25% KV-cache headroom ≈ 149 GB
- 2× H200 141GB at $3.59–$7.91/GPU-hour on-demand (2026-09-24)
- ~8,963 output tok/s aggregate at batch 322 (bandwidth-bound); MoE compute scales with active params
- Blended 3:1 input:output; prefill ~143,858 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)~69.1%
- #1Claude Opus 596.0%
- #2Claude Sonnet 585.2%
- #16Nemotron 3 Ultra70.7%
- ~#17Mistral Small 4Estimated69.1%
- #17Intern-S2-Preview (35B-A3B)64.0%
No published score — estimated from related benchmarks
SWE-bench Pro (Public)Estimated
~#23 of 24 (estimated)~47.2%
- #1Claude Fable 5.293.0%
- #2Claude Opus 5.589.9%
- #22Muse Glimmer51.2%
- ~#23Mistral Small 4Estimated47.2%
- #23Nex-N2.5-Mini43.8%
No published score — estimated from related benchmarks
DeepSWE v1.1Estimated
~#41 of 46 (estimated)~40.7%
- #1Gemini 4 Argon78.8%
- #2DeepSeek V4.1 Flash 101777.0%
- #40Qwen3.8 27B42.2%
- ~#41Mistral Small 4Estimated40.7%
- #41Laguna S 2.140.4%
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%
- #63Mistral Medium 3.50.0%
- #63Mistral Small 40.0%
- #63Nemotron 3 Super 120B A12B0.0%
artificialanalysis.ai · 2026-09-24
LiveCodeBenchLab-reported
#15 of 2764.0%
- #1Qwen3.8-Omni-Flash92.6%
- #2Qwen3.8 Flash-Next91.9%
- #14Maple-Preview75.1%
- #15Mistral Small 464.0%
- #16Claude Haiku 4.561.5%
huggingface.co · 2026-03-16
SciCodeVerified
#86 of 11838.8%
- #1Claude Fable 5.267.7%
- #2Claude Opus 5.566.9%
- #85gpt-oss-20b38.9%
- #86Mistral Small 438.8%
- #86North Mini Code38.8%
artificialanalysis.ai · 2026-09-24
AA Coding Agent IndexEstimated
~#25 of 26 (estimated)~27.9
- #1Claude Opus 5.561.7
- #2Claude Opus 559.7
- #24GPT-6 Luna31.1
- ~#25Mistral Small 4Estimated27.9
- #25Muse Spark Open26.3
No published score — estimated from related benchmarks
Writing & chat
IFBenchVerified
#75 of 13948.2%
- #1MiniMax-M382.9%
- #2Qwen3.8 2.4T A95B82.8%
- #74MiniCPM5-1B49.3%
- #75Mistral Small 448.2%
- #76Llama 3.3 Instruct 70B47.1%
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
~#45 of 57 (estimated)~1360 Elo
- #1Gemini 4 ArgonFew votes1512
- #2Claude Opus 5.51497
- #44GLM-4.6VFew votes1361
- ~#45Mistral Small 4Estimated1360
- #45Trinity Large Thinking1350
No published score — estimated from related benchmarks
LMArena Text (Elo)Estimated
~#44 of 57 (estimated)~1380 Elo
- #1Gemini 4 Argon1525
- #2Claude Opus 5.51507
- #43Qwen3 Next 80B A3B Instruct1399
- ~#44Mistral Small 4Estimated1380
- #44GLM-4.6VFew votes1378
No published score — estimated from related benchmarks
EQ-Bench Creative Writing v3 (Elo)Estimated
~#24 of 26 (estimated)~1324 Elo
- #1GPT-6 Astra2173
- #2Claude Fable 5.12162
- #23Gemma 4 31B1368
- ~#24Mistral Small 4Estimated1324
- #24gpt-oss-120b961
No published score — estimated from related benchmarks
Vision
MMMU-ProVerified
#65 of 8756.8%
- #1Claude Opus 5.587.7%
- #2GPT-6 Astra86.9%
- #64Claude Haiku 4.558.6%
- #65Mistral Small 456.8%
- #66Mistral Large 355.7%
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%
- ~#12Mistral Small 4Estimated<74.5%
No published score — estimated from related benchmarks
ChartographyEstimated
~last of 25 (estimated)below measured range (<9.0%)
No published score — estimated from related benchmarks
Hard reasoning
GPQA DiamondVerified
#85 of 19276.9%
- #1GPT-6 Astra96.1%
- #2Gemini 3.8 Flash95.3%
- #84LongCat 2.078.0%
- #85Mistral Small 476.9%
- #86Cogito v2.176.8%
artificialanalysis.ai · 2026-09-24
Humanity's Last ExamVerified
#121 of 1989.9%
- #1Claude Fable 5.274.3%
- #2Claude Opus 5.561.4%
- #119Qwen3 Coder Next10.1%
- #121Mistral Small 49.9%
- #122Granite 4.2 8B9.7%
artificialanalysis.ai · 2026-09-24
AIME (latest)Lab-reported
#19 of 3384.0%
- #1GLM-5.3 NVFP499.5%
- #2Beam 1.0 501B A23B97.7%
- #18Mistral Medium 3.586.3%
- #19Mistral Small 484.0%
- #20Claude Haiku 4.583.7%
huggingface.co · 2026-03-16
Agents
τ²-bench (Telecom)Verified
#55 of 11941.2%
- #1Step 3.7 Flash98.5%
- #2DeepSeek-V4-Pro-0813-nvfp4-DSpark98.3%
- #54Gemma 4 26B A4B43.6%
- #55Mistral Small 441.2%
- #56NVIDIA Nemotron 3 Nano 30B A3B40.9%
artificialanalysis.ai · 2026-09-24
OSWorld-VerifiedEstimated
~#12 of 15 (estimated)~56.2%
- #1Kimi K384.8%
- #2Qwen3.8 27B84.3%
- #11Muse Glimmer65.9%
- ~#12Mistral Small 4Estimated56.2%
- #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
GDPval-AA (Elo)Verified
#80 of 91349 Elo
- #1Claude Opus 5.51866
- #2Muse Wonder Full1857
- #79Nova 2.0 Lite354
- #80Mistral Small 4349
- #81gpt-oss-20b327
artificialanalysis.ai · 2026-09-24