Model comparison

Mistral Medium 3.5 vs Step 5 Preview

Step 5 Preview is the stronger model overall, scoring 43.8 to 39.6 on the Noometry Index.

Last verified . 1 shared benchmarks.

Mistral Medium 3.5 Mistral AI

39.6

Rank #151 Confirmed

Step 5 Preview StepFun

43.8

Rank #78 Reported

Summary

  • They share 1 benchmark with published results for both. Mistral Medium 3.5 scores higher in 0 categories and Step 5 Preview in 2 categories; 2 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where Step 5 Preview leads 38.0 to 16.6.
  • Step 5 Preview is cheaper at $1 / $2.70 per million input/output tokens, against $1.50 / $7.50 for Mistral Medium 3.5.
  • Step 5 Preview accepts more context: 1.02M tokens versus 262K.
  • Mistral Medium 3.5 has downloadable open weights; the other is API-only.

Side by side

Mistral Medium 3.5 and Step 5 Preview specifications
Mistral Medium 3.5Step 5 Preview
ProviderMistral AIStepFun
Noometry Index39.643.8
Released—2026-09-16
WeightsOpenProprietary
Context window262K1.02M
Max output210K66K
Input $ / M tokens$1.50$1
Output $ / M tokens$7.50$2.70
Results tracked223

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

Category by category

Coding Step 5 Preview leads

Mistral Medium 3.5: 35.7 (#208), Step 5 Preview: 50.6 (#40)

Coding benchmarks
BenchmarkMistral Medium 3.5Step 5 Preview
LMArena WebDev12631569
SciCode—58.9%
LMArena Coding1461—

Reasoning Step 5 Preview leads

Mistral Medium 3.5: 16.6 (#290), Step 5 Preview: 38.0 (#58)

Reasoning benchmarks
BenchmarkMistral Medium 3.5Step 5 Preview
Kagi LLM Benchmark41.4%—
NYT Connections (extended)12.9%—
CritPt—20.9%
LMArena Hard Prompts1436—
Epoch Capabilities Index141.42—

Math Not comparable

Mistral Medium 3.5: 36.4 (#113), Step 5 Preview: —

Math benchmarks
BenchmarkMistral Medium 3.5Step 5 Preview
LMArena Math1432—

Knowledge Not comparable

Mistral Medium 3.5: 39.3 (#123), Step 5 Preview: —

Knowledge benchmarks
BenchmarkMistral Medium 3.5Step 5 Preview
LMArena Expert1431—

Multimodal Not comparable

Mistral Medium 3.5: 37.6 (#64), Step 5 Preview: —

Multimodal benchmarks
BenchmarkMistral Medium 3.5Step 5 Preview
LMArena Vision1222—

Multilingual Not comparable

Mistral Medium 3.5: 52.1 (#93), Step 5 Preview: —

Multilingual benchmarks
BenchmarkMistral Medium 3.5Step 5 Preview
LMArena Non-English1405—
LMArena Chinese1443—
LMArena French1448—
LMArena German1453—
LMArena Korean1387—
LMArena Russian1396—
LMArena Spanish1406—

Instruction Following Not comparable

Mistral Medium 3.5: 74.7 (#87), Step 5 Preview: —

Instruction Following benchmarks
BenchmarkMistral Medium 3.5Step 5 Preview
LMArena Instruction Following1414—

Long Context Not comparable

Mistral Medium 3.5: 43.2 (#101), Step 5 Preview: —

Long Context benchmarks
BenchmarkMistral Medium 3.5Step 5 Preview
LMArena Longer Query1414—

Writing & Preference Not comparable

Mistral Medium 3.5: 58.6 (#113), Step 5 Preview: —

Writing & Preference benchmarks
BenchmarkMistral Medium 3.5Step 5 Preview
LMArena Text1421—
LMArena Creative Writing1372—
EQ-Bench 4993—
LMArena Multi-Turn1423—

Frequently asked questions

Is Mistral Medium 3.5 better than Step 5 Preview?

Step 5 Preview is the stronger model overall, scoring 43.8 to 39.6 on the Noometry Index.

Which is cheaper, Mistral Medium 3.5 or Step 5 Preview?

Step 5 Preview is cheaper. It lists at $1 per million input tokens and $2.70 per million output tokens; Mistral Medium 3.5 lists at $1.50 and $7.50.

Is Mistral Medium 3.5 or Step 5 Preview better for coding?

Step 5 Preview scores higher on coding benchmarks: 50.6 versus 35.7 in the Noometry coding category.

Which has the bigger context window?

Step 5 Preview does, with 1.02M tokens against 262K.

How many benchmarks do Mistral Medium 3.5 and Step 5 Preview share?

1 benchmark has published results for both models. Mistral Medium 3.5 has 22 scored results on Noometry and Step 5 Preview has 3.

Related comparisons

Go deeper