Noometry

StepFun, proprietary

Step 5 Preview

Step 5 Preview by StepFun ranks 58th of 354 ranked models on the Noometry Index as of October 2026, with a score of 47.9. Its strongest category is coding, where it ranks 36th. API pricing starts at $1 per million input tokens and $2.70 per million output tokens, with a 1.02M-token context window.

Last verified

Noometry rank
#58 of 354
Index score
47.9
Evidence
Confirmed 18 results
Provider
StepFun
Released
September 16, 2026
Weights
Proprietary
Reasoning
Yes
Context window
1.02M
Max output
66K
Input price
$1 / M
Output price
$2.70 / M
Blended price
$1.43 / M
Output speed
Not measured
Value
#129 of 219
Knowledge cutoff
Unknown
Input
text, image, video

Category scores

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

  1. Coding 51.8
  2. Reasoning 40.0
  3. Math 46.1
  4. Knowledge 41.2
  5. Multimodal 41.0
  6. Multilingual 53.8
  7. Instruction Following 76.0
  8. Long Context 44.5
  9. Writing & Preference 62.8
CategoryScoreRankResults
Coding51.8#363
Reasoning40.0#572
Math46.1#722
Knowledge41.2#1121
Multimodal41.0#411
Multilingual53.8#541
Instruction Following76.0#501
Long Context44.5#641
Writing & Preference62.8#713

Strengths and weaknesses

Categories where Step 5 Preview places highest and lowest among the models ranked in each, with its score against that category's median.

Strongest categories

CategoryScorevs medianRank
Coding51.8+13.1#36 of 340, top 11%
Reasoning40.0+16.4#57 of 350, top 17%
Instruction Following76.0+4.7#50 of 305, top 17%

Weakest categories

CategoryScorevs medianRank
Knowledge41.2+3.8#112 of 314, top 36%
Multimodal41.0+2.5#41 of 128, top 33%
Writing & Preference62.8+9.0#71 of 312, top 23%

Closest competitors

The models ranked just above and below Step 5 Preview. When scores are this close, price and speed are often the better way to choose.

ModelRankScoreBlended $/MSpeed
Grok 4.20 (Non-Reasoning)#5448.6$1.5661Compare
MiMo-V2.6-Flash#5548.5$0.18—Compare
Grok 4#5648.1—1Compare
Kimi K2.5#5748.1$0.9066Compare
GLM-5.1#5947.8$2.15—Compare
Kimi K2.6#6047.7$1.71—Compare
o3#6147.5$3.503Compare
Qwen3.6 Plus#6247.5$1.13—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
LMArena WebDev1564#30 of 113, top 27%highLMArena2026-10-08
SciCode58.9%#12 of 121, top 10%Epoch AI
LMArena Coding1480#47 of 294, top 16%Step 5 PreviewLMArena2026-10-08

Reasoning

BenchmarkScorePositionSettingSourceDate
CritPt20.9%#22 of 134, top 17%Epoch AI
LMArena Hard Prompts1465#42 of 297, top 15%Step 5 PreviewLMArena2026-10-08

Math

BenchmarkScorePositionSettingSourceDate
ProofBench42%#35 of 77, top 46%Epoch AI
LMArena Math1470#41 of 285, top 15%Step 5 PreviewLMArena2026-10-08

Knowledge

BenchmarkScorePositionSettingSourceDate
LMArena Expert1470#50 of 273, top 19%Step 5 PreviewLMArena2026-10-08

Multimodal

BenchmarkScorePositionSettingSourceDate
LMArena Vision1267#40 of 122, top 33%Step 5 PreviewLMArena2026-10-09

Multilingual

BenchmarkScorePositionSettingSourceDate
LMArena Non-English1432#54 of 297, top 19%Step 5 PreviewLMArena2026-10-08
LMArena Chinese1519#26 of 285, top 10%Step 5 PreviewLMArena2026-10-08
LMArena Russian1436#58 of 283, top 21%Step 5 PreviewLMArena2026-10-08
LMArena Spanish1439#68 of 226, top 31%Step 5 PreviewLMArena2026-10-08

Instruction Following

BenchmarkScorePositionSettingSourceDate
LMArena Instruction Following1444#46 of 298, top 16%Step 5 PreviewLMArena2026-10-08

Long Context

BenchmarkScorePositionSettingSourceDate
LMArena Longer Query1455#45 of 291, top 16%Step 5 PreviewLMArena2026-10-08

Writing & Preference

BenchmarkScorePositionSettingSourceDate
LMArena Text1442#58 of 297, top 20%Step 5 PreviewLMArena2026-10-08
LMArena Creative Writing1410#64 of 295, top 22%Step 5 PreviewLMArena2026-10-08
LMArena Multi-Turn1447#59 of 295, top 20%Step 5 PreviewLMArena2026-10-08

API pricing by provider

RouteInput $/MOutput $/MCached input $/MChecked
openrouter$1$2.70$0.052026-10-10
stepfun$1$2.70$0.052026-10-10

All StepFun API prices → Estimate your cost →

Compare Step 5 Preview

Other StepFun models

Frequently asked questions

How good is Step 5 Preview?

Step 5 Preview by StepFun ranks 58th of 354 ranked models on the Noometry Index as of October 2026, with a score of 47.9. Its strongest category is coding, where it ranks 36th. API pricing starts at $1 per million input tokens and $2.70 per million output tokens, with a 1.02M-token context window.

How much does Step 5 Preview cost?

Step 5 Preview costs $1 per million input tokens and $2.70 per million output tokens on StepFun's own API, with cached input at $0.05.

What is Step 5 Preview's context window?

Step 5 Preview accepts up to 1.02M tokens of input and can write up to 66K tokens in one response.

Is Step 5 Preview open source?

No. Step 5 Preview is proprietary and available only through StepFun's API and partner platforms.

What are Step 5 Preview's strengths and weaknesses?

Relative to other ranked models, Step 5 Preview places best in coding, reasoning, instruction following and lowest in knowledge, multimodal, writing & preference.

What is Step 5 Preview best at?

Its best category is coding, where it ranks 36th on Noometry.