Noometry

Mistral AI, open weights

Mistral Medium

Mistral Medium by Mistral AI ranks 218th of 354 ranked models on the Noometry Index as of October 2026, with a score of 36.3. Its strongest category is multimodal, where it ranks 88th. API pricing starts at $1.50 per million input tokens and $7.50 per million output tokens, with a 262K-token context window.

Last verified

Noometry rank
#218 of 354
Index score
36.3
Evidence
Confirmed 36 results
Provider
Mistral AI
Released
December 11, 2023
Weights
Open weights
Reasoning
Yes
Context window
262K
Max output
262K
Input price
$1.50 / M
Output price
$7.50 / M
Blended price
$3 / M
Output speed
68 tokens/s Kagi
Value
#177 of 219
Knowledge cutoff
May 2025
Input
text, image

Category scores

Each category score combines every public result we have in that category.

  1. Coding 34.2
  2. Agentic & Tool Use 28.3
  3. Reasoning 24.0
  4. Math 28.1
  5. Knowledge 25.0
  6. Multimodal 35.3
  7. Multilingual 52.1
  8. Instruction Following 73.7
  9. Long Context 42.9
  10. Writing & Preference 60.0
CategoryScoreRankResults
Coding34.2#2434
Agentic & Tool Use28.3#901
Reasoning24.0#1676
Math28.1#2454
Knowledge25.0#2654
Multimodal35.3#881
Multilingual52.1#911
Instruction Following73.7#1161
Long Context42.9#1141
Writing & Preference60.0#1034

Strengths and weaknesses

Categories where Mistral Medium places highest and lowest among the models ranked in each, with its score against that category's median.

Strongest categories

CategoryScorevs medianRank
Multilingual52.1+4.7#91 of 297, top 31%
Writing & Preference60.0+6.2#103 of 312, top 34%
Instruction Following73.7+2.5#116 of 305, top 39%

Weakest categories

CategoryScorevs medianRank
Knowledge25.0−12.3#265 of 314, top 85%
Math28.1−8.5#245 of 327, top 75%
Coding34.2−4.5#243 of 340, top 72%

Closest competitors

The models ranked just above and below Mistral Medium. When scores are this close, price and speed are often the better way to choose.

ModelRankScoreBlended $/MSpeed
Granite 4.0 H Small#21436.5——Compare
Command A#21536.5$4.3828Compare
Grok Build 0.1#21636.4$1.25—Compare
gpt-oss-120b#21736.3$0.070355Compare
GPT-4.1#21935.9$3.50116Compare
Deepseek Coder v2#22035.9——Compare
C4ai Aya Expanse 32b#22135.9——Compare
Llama 3.1 Nemotron 51b Instruct#22235.9——Compare

Sponsored placements are available on pages like this one. Advertise on Noometry

Benchmark results

Every published result we track, with its source. Bold rows are the ones used for ranking; where several exist we prefer independent runs over self-reported numbers.

Coding

BenchmarkScorePositionSettingSourceDate
FrontierCode8%#37 of 37, top 100%Epoch AI
SciCode33.8%Epoch AI
SciCode40.2%#78 of 121, top 65%Epoch AI
WeirdML33.1%Epoch AI
WeirdML43.7%#63 of 119, top 53%Epoch AI
LMArena Coding1434#106 of 294, top 37%LMArena2026-10-08
LMArena Coding1386LMArena2026-10-08
ALE-Bench763.98#65 of 105, top 62%Epoch AI
ALE-Bench210.18Epoch AI

Agentic & Tool Use

BenchmarkScorePositionSettingSourceDate
Berkeley Function Calling Leaderboard37.7%#25 of 49, top 52%Berkeley Function Calling Leaderboard

Reasoning

BenchmarkScorePositionSettingSourceDate
Kagi LLM Benchmark50%#64 of 99, top 65%Kagi LLM Benchmark
CritPt0%#124 of 134, top 93%Epoch AI
CritPt0%#124 of 134, top 93%Epoch AI
LMArena Hard Prompts1426#97 of 297, top 33%LMArena2026-10-08
LMArena Hard Prompts1365LMArena2026-10-08
DTBench53.3%Epoch AI
DTBench62.3%Epoch AI
DTBench67.5%Epoch AI
DTBench75.5%#86 of 151, top 57%Epoch AI
LMCA19.1%Epoch AI
LMCA26.1%#84 of 125, top 68%Epoch AI
Surface Evolver Bench26.9%#22 of 25, top 88%Epoch AI

Math

BenchmarkScorePositionSettingSourceDate
OTIS Mock AIME 2024-202532.2%#119 of 173, top 69%Epoch AI2025-05-07
ProofBench9%#65 of 77, top 85%Epoch AI
LMArena Math1408#114 of 285, top 40%LMArena2026-10-08
LMArena Math1353LMArena2026-10-08
MATH Level 581.6%#25 of 79, top 32%Epoch AI2025-05-07
FrontierMath (Feb 2025 set)0.3%#63 of 68, top 93%Epoch AI2025-05-07

Knowledge

BenchmarkScorePositionSettingSourceDate
GPQA Diamond59.5%#115 of 186, top 62%Epoch AI2025-05-07
Humanity's Last Exam4.5%#37 of 41, top 91%Epoch AI
Vectara Hallucination Rate (lower is better)22.7%#94 of 96, top 98%Vectara Hallucination Leaderboard
LMArena Expert1342LMArena2026-10-08
LMArena Expert1408#115 of 273, top 43%LMArena2026-10-08

Multimodal

BenchmarkScorePositionSettingSourceDate
LMArena Vision1157LMArena2026-10-09
LMArena Vision1172#90 of 122, top 74%LMArena2026-10-09

Multilingual

BenchmarkScorePositionSettingSourceDate
LMArena Non-English1353LMArena2026-10-08
LMArena Non-English1408#91 of 297, top 31%LMArena2026-10-08
LMArena Chinese1374LMArena2026-10-08
LMArena Chinese1447#104 of 285, top 37%LMArena2026-10-08
LMArena French1459#52 of 223, top 24%LMArena2026-10-08
LMArena French1368LMArena2026-10-08
LMArena German1432#63 of 231, top 28%LMArena2026-10-08
LMArena German1377LMArena2026-10-08
LMArena Japanese1316LMArena2026-10-08
LMArena Japanese1378#80 of 211, top 38%LMArena2026-10-08
LMArena Korean1304LMArena2026-10-08
LMArena Korean1380#77 of 213, top 37%LMArena2026-10-08
LMArena Russian1359LMArena2026-10-08
LMArena Russian1411#93 of 283, top 33%LMArena2026-10-08
LMArena Spanish1433#77 of 226, top 35%LMArena2026-10-08
LMArena Spanish1364LMArena2026-10-08

Instruction Following

BenchmarkScorePositionSettingSourceDate
LMArena Instruction Following1339LMArena2026-10-08
LMArena Instruction Following1398#109 of 298, top 37%LMArena2026-10-08

Long Context

BenchmarkScorePositionSettingSourceDate
LMArena Longer Query1406#111 of 291, top 39%LMArena2026-10-08
LMArena Longer Query1360LMArena2026-10-08

Writing & Preference

BenchmarkScorePositionSettingSourceDate
LMArena Text1424#88 of 297, top 30%LMArena2026-10-08
LMArena Text1370LMArena2026-10-08
LMArena Creative Writing1344LMArena2026-10-08
LMArena Creative Writing1391#94 of 295, top 32%LMArena2026-10-08
Short-Story Creative Writing77.3%#17 of 39, top 44%Epoch AI
LMArena Multi-Turn1418#97 of 295, top 33%LMArena2026-10-08
LMArena Multi-Turn1383LMArena2026-10-08

API pricing by provider

RouteInput $/MOutput $/MCached input $/MChecked
azure$0.40$2—2026-10-10
mistral$1.50$7.50$0.152026-10-10

All Mistral AI API prices → Estimate your cost →

Compare Mistral Medium

Other Mistral AI models

Frequently asked questions

How good is Mistral Medium?

Mistral Medium by Mistral AI ranks 218th of 354 ranked models on the Noometry Index as of October 2026, with a score of 36.3. Its strongest category is multimodal, where it ranks 88th. API pricing starts at $1.50 per million input tokens and $7.50 per million output tokens, with a 262K-token context window.

How much does Mistral Medium cost?

Mistral Medium costs $1.50 per million input tokens and $7.50 per million output tokens on Mistral AI's own API, with cached input at $0.15.

What is Mistral Medium's context window?

Mistral Medium accepts up to 262K tokens of input and can write up to 262K tokens in one response.

Is Mistral Medium open source?

Yes. Mistral Medium's weights are downloadable; check the license for commercial terms.

How fast is Mistral Medium?

Mistral Medium generated about 68 output tokens per second in the Kagi LLM Benchmark's timed runs. Speed varies by provider, load and reasoning effort.

What are Mistral Medium's strengths and weaknesses?

Relative to other ranked models, Mistral Medium places best in multilingual, writing & preference, instruction following and lowest in knowledge, math, coding.

What is Mistral Medium best at?

Its best category is multimodal, where it ranks 88th on Noometry.