Model comparison

DeepSeek-V3.1-Terminus vs GLM-5.2

GLM-5.2 is the stronger model overall, scoring 50.9 to 42.6 on the Noometry Index. DeepSeek-V3.1-Terminus costs 4.5× less per token, which makes it the better buy when GLM-5.2's lead doesn't matter for your workload.

Last verified October 10, 2026. 16 shared benchmarks.

DeepSeek-V3.1-Terminus

DeepSeek

42.6

Rank #96 Confirmed

GLM-5.2 Z.ai (Zhipu)

50.9

Rank #38 Confirmed

Summary

They share 16 benchmarks with published results for both. DeepSeek-V3.1-Terminus scores higher in 0 categories and GLM-5.2 in 7 categories; 7 gaps are clear of the uncertainty.
The widest gap is in math, where GLM-5.2 leads 55.4 to 35.7.
The biggest single-benchmark swing is CritPt: 1.7% for DeepSeek-V3.1-Terminus and 20.9% for GLM-5.2.
DeepSeek-V3.1-Terminus is cheaper at $0.30 / $1 per million input/output tokens, against $1.40 / $4.40 for GLM-5.2.
GLM-5.2 accepts more context: 1M tokens versus 164K.

Side by side

DeepSeek-V3.1-Terminus and GLM-5.2 specifications
	DeepSeek-V3.1-Terminus	GLM-5.2
Provider	DeepSeek	Z.ai (Zhipu)
Noometry Index	42.6	50.9
Released	2025-09-22	2026-06-13
Weights	Open	Open
Context window	164K	1M
Max output	66K	131K
Input $ / M tokens	$0.30	$1.40
Output $ / M tokens	$1	$4.40
Results tracked	16	50

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

Category by category

Coding GLM-5.2 leads

DeepSeek-V3.1-Terminus: 41.5 (#114), GLM-5.2: 50.9 (#37)

Coding benchmarks
Benchmark	DeepSeek-V3.1-Terminus	GLM-5.2
SciCode	40.6%	50.5%
LMArena Coding	1426	1484
ALE-Bench	745.17	1,047
SWE-bench Verified	—	78.7%
DeepSWE	—	43.8%
FrontierCode	—	24.5%
LMArena WebDev	—	1604
WeirdML	—	70.1%

Agentic & Tool Use Not comparable

DeepSeek-V3.1-Terminus: —, GLM-5.2: 31.6 (#63)

Agentic & Tool Use benchmarks
Benchmark	DeepSeek-V3.1-Terminus	GLM-5.2
τ²-bench Banking	—	37.1%
PostTrainBench	—	31.7%
GBAEval	—	0%
Vending-Bench 2	—	8,314

Reasoning GLM-5.2 leads

DeepSeek-V3.1-Terminus: 25.7 (#126), GLM-5.2: 42.1 (#46)

Reasoning benchmarks
Benchmark	DeepSeek-V3.1-Terminus	GLM-5.2
Kagi LLM Benchmark	57.4%	62.6%
CritPt	1.7%	20.9%
LMArena Hard Prompts	1427	1480
DTBench	81.3%	93.6%
LMCA	28.6%	45.8%
ARC-AGI-2	—	22.8%
SimpleBench	—	58.8%
NYT Connections (extended)	—	74.3%
ARC-AGI-1	—	77%
Chess Puzzles	—	21%
EBR-Bench	—	9.5%
Mystery Game Puzzles	—	19%
Surface Evolver Bench	—	55.6%
Epoch Capabilities Index	—	151.76

Math GLM-5.2 leads

DeepSeek-V3.1-Terminus: 35.7 (#138), GLM-5.2: 55.4 (#35)

Math benchmarks
Benchmark	DeepSeek-V3.1-Terminus	GLM-5.2
LMArena Math	1402	1482
FrontierMath (Tiers 1-3)	—	59.2%
FrontierMath Tier 4	—	29.3%
MathArena Final-Answer Competitions	—	67.6%
OTIS Mock AIME 2024-2025	—	86.4%
ProofBench	—	35%

Knowledge Not comparable

DeepSeek-V3.1-Terminus: —, GLM-5.2: 57.1 (#36)

Knowledge benchmarks
Benchmark	DeepSeek-V3.1-Terminus	GLM-5.2
GPQA Diamond	—	91.9%
SimpleQA Verified	—	34.2%
LMArena Expert	—	1484

Multilingual GLM-5.2 leads

DeepSeek-V3.1-Terminus: 52.3 (#87), GLM-5.2: 56.0 (#25)

Multilingual benchmarks
Benchmark	DeepSeek-V3.1-Terminus	GLM-5.2
LMArena Non-English	1407	1459
LMArena Russian	1434	1467
LMArena Chinese	—	1518
LMArena French	—	1476
LMArena German	—	1470
LMArena Japanese	—	1451
LMArena Korean	—	1445
LMArena Spanish	—	1477

Instruction Following GLM-5.2 leads

DeepSeek-V3.1-Terminus: 74.4 (#96), GLM-5.2: 77.1 (#32)

Instruction Following benchmarks
Benchmark	DeepSeek-V3.1-Terminus	GLM-5.2
LMArena Instruction Following	1406	1465

Long Context GLM-5.2 leads

DeepSeek-V3.1-Terminus: 43.5 (#88), GLM-5.2: 45.3 (#43)

Long Context benchmarks
Benchmark	DeepSeek-V3.1-Terminus	GLM-5.2
LMArena Longer Query	1426	1479

Writing & Preference GLM-5.2 leads

DeepSeek-V3.1-Terminus: 61.3 (#87), GLM-5.2: 70.7 (#20)

Writing & Preference benchmarks
Benchmark	DeepSeek-V3.1-Terminus	GLM-5.2
LMArena Text	1420	1470
LMArena Creative Writing	1404	1465
LMArena Multi-Turn	1411	1468
EQ-Bench Creative Writing	—	1757
EQ-Bench 4	—	1222

Frequently asked questions

Is DeepSeek-V3.1-Terminus better than GLM-5.2?

Which is cheaper, DeepSeek-V3.1-Terminus or GLM-5.2?

DeepSeek-V3.1-Terminus is cheaper. It lists at $0.30 per million input tokens and $1 per million output tokens; GLM-5.2 lists at $1.40 and $4.40.

Is DeepSeek-V3.1-Terminus or GLM-5.2 better for coding?

GLM-5.2 scores higher on coding benchmarks: 50.9 versus 41.5 in the Noometry coding category.

Which has the bigger context window?

GLM-5.2 does, with 1M tokens against 164K.

How many benchmarks do DeepSeek-V3.1-Terminus and GLM-5.2 share?

16 benchmarks have published results for both models. DeepSeek-V3.1-Terminus has 16 scored results on Noometry and GLM-5.2 has 50.