Model comparison

DeepSeek-V3.1-Terminus vs Step 5 Preview

Step 5 Preview is the stronger model overall, scoring 47.9 to 43.1 on the Noometry Index. DeepSeek-V3.1-Terminus costs 3.1× less per token, which makes it the better buy when Step 5 Preview's lead doesn't matter for your workload.

Last verified October 10, 2026. 12 shared benchmarks.

DeepSeek-V3.1-Terminus

DeepSeek

43.1

Rank #97 Confirmed

Step 5 Preview

StepFun

47.9

Rank #58 Confirmed

Summary

They share 12 benchmarks with published results for both. DeepSeek-V3.1-Terminus scores higher in 0 categories and Step 5 Preview in 7 categories; 7 gaps are clear of the uncertainty.
The widest gap is in reasoning, where Step 5 Preview leads 40.0 to 26.4.
The biggest single-benchmark swing is CritPt: 1.7% for DeepSeek-V3.1-Terminus and 20.9% for Step 5 Preview.
DeepSeek-V3.1-Terminus is cheaper at $0.27 / $1 per million input/output tokens, against $1 / $2.70 for Step 5 Preview.
Step 5 Preview accepts more context: 1.02M tokens versus 164K.
DeepSeek-V3.1-Terminus has downloadable open weights; the other is API-only.

Side by side

DeepSeek-V3.1-Terminus and Step 5 Preview specifications
	DeepSeek-V3.1-Terminus	Step 5 Preview
Provider	DeepSeek	StepFun
Noometry Index	43.1	47.9
Released	2025-09-22	2026-09-16
Weights	Open	Proprietary
Context window	164K	1.02M
Max output	147K	66K
Input $ / M tokens	$0.27	$1
Output $ / M tokens	$1	$2.70
Results tracked	16	18

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

Category by category

Coding Step 5 Preview leads

DeepSeek-V3.1-Terminus: 42.0 (#113), Step 5 Preview: 51.8 (#36)

Coding benchmarks
Benchmark	DeepSeek-V3.1-Terminus	Step 5 Preview
SciCode	40.6%	58.9%
LMArena Coding	1426	1480
LMArena WebDev	—	1564
ALE-Bench	745.17	—

Reasoning Step 5 Preview leads

DeepSeek-V3.1-Terminus: 26.4 (#133), Step 5 Preview: 40.0 (#57)

Reasoning benchmarks
Benchmark	DeepSeek-V3.1-Terminus	Step 5 Preview
CritPt	1.7%	20.9%
LMArena Hard Prompts	1426	1465
Kagi LLM Benchmark	57.4%	—
DTBench	81.3%	—
LMCA	28.6%	—

Math Step 5 Preview leads

DeepSeek-V3.1-Terminus: 38.5 (#137), Step 5 Preview: 46.1 (#72)

Math benchmarks
Benchmark	DeepSeek-V3.1-Terminus	Step 5 Preview
LMArena Math	1402	1470
ProofBench	—	42%

Knowledge Not comparable

DeepSeek-V3.1-Terminus: —, Step 5 Preview: 41.2 (#112)

Knowledge benchmarks
Benchmark	DeepSeek-V3.1-Terminus	Step 5 Preview
LMArena Expert	—	1470

Multimodal Not comparable

DeepSeek-V3.1-Terminus: —, Step 5 Preview: 41.0 (#41)

Multimodal benchmarks
Benchmark	DeepSeek-V3.1-Terminus	Step 5 Preview
LMArena Vision	—	1267

Multilingual Step 5 Preview leads

DeepSeek-V3.1-Terminus: 52.1 (#92), Step 5 Preview: 53.8 (#54)

Multilingual benchmarks
Benchmark	DeepSeek-V3.1-Terminus	Step 5 Preview
LMArena Non-English	1407	1432
LMArena Russian	1436	1436
LMArena Chinese	—	1519
LMArena Spanish	—	1439

Instruction Following Step 5 Preview leads

DeepSeek-V3.1-Terminus: 74.0 (#106), Step 5 Preview: 76.0 (#50)

Instruction Following benchmarks
Benchmark	DeepSeek-V3.1-Terminus	Step 5 Preview
LMArena Instruction Following	1404	1444

Long Context Step 5 Preview leads

DeepSeek-V3.1-Terminus: 43.4 (#97), Step 5 Preview: 44.5 (#64)

Long Context benchmarks
Benchmark	DeepSeek-V3.1-Terminus	Step 5 Preview
LMArena Longer Query	1421	1455

Writing & Preference Step 5 Preview leads

DeepSeek-V3.1-Terminus: 61.0 (#92), Step 5 Preview: 62.8 (#71)

Writing & Preference benchmarks
Benchmark	DeepSeek-V3.1-Terminus	Step 5 Preview
LMArena Text	1419	1442
LMArena Creative Writing	1403	1410
LMArena Multi-Turn	1411	1447

Frequently asked questions

Is DeepSeek-V3.1-Terminus better than Step 5 Preview?

Which is cheaper, DeepSeek-V3.1-Terminus or Step 5 Preview?

DeepSeek-V3.1-Terminus is cheaper. It lists at $0.27 per million input tokens and $1 per million output tokens; Step 5 Preview lists at $1 and $2.70.

Is DeepSeek-V3.1-Terminus or Step 5 Preview better for coding?

Step 5 Preview scores higher on coding benchmarks: 51.8 versus 42.0 in the Noometry coding category.

Which has the bigger context window?

Step 5 Preview does, with 1.02M tokens against 164K.

How many benchmarks do DeepSeek-V3.1-Terminus and Step 5 Preview share?

12 benchmarks have published results for both models. DeepSeek-V3.1-Terminus has 16 scored results on Noometry and Step 5 Preview has 18.