DeepSeek V3.2 vs GLM 5.1

GLM 5.1 leads DeepSeek V3.2 on coding, 49.2 to 30.0 on the Vibe Coding Index, and DeepSeek V3.2 costs 5.0x less per output token ($0.88 vs $4.40 per million). Updated October 21, 2026; 60 benchmarks have a result for both.

Replace DeepSeek V3.2 Replace GLM 5.1

DeepSeek V3.2DeepSeek

Coding: 30.0
Overall: 37.9
Output $/M: $0.88

GLM 5.1Z.ai

Coding: 49.2
Overall: 57.5
Output $/M: $4.40

vs boards won of 60 shared

DeepSeek V3.2 or GLM 5.1: which to pick

Pick GLM 5.1 for coding: it scores 49.2 on the Vibe Coding Index against 30.0 for DeepSeek V3.2.

Pick DeepSeek V3.2 for output tokens 5.0x cheaper ($0.88 vs $4.40 per million) and work measured by CaseLaw (Vals AI), where it leads.

Model ADeepSeek V3.2DeepSeek30.0/1003 independent benchmarks$0.99 per solved task Model BGLM 5.1Z.ai49.2/1003 independent benchmarks$0.60 per solved task

601529

Shared results: 60
DeepSeek V3.2 only: 15
GLM 5.1 only: 29
Like-for-like categories: 8 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

Coding
Vibe Coding Index, six public coding benchmarks
GLM 5.1
GLM 5.1 scores 49.2 against 30.0.
3 benchmarks behind each score
Intelligence
AA Intelligence Index
GLM 5.1
GLM 5.1 leads, 26.1 against 21.5.
Same board, both measured
Agentic work
AA Agentic Index
No call
Only GLM 5.1 has a AA Agentic Index result, 23.9.
Directional only
Reasoning
GPQA Diamond
GLM 5.1
GLM 5.1 leads, 84.5% against 80.3%.
Same board, both measured
Math
AIME
GLM 5.1
GLM 5.1 leads, 91.9% against 84.6%.
Same board, both measured
Knowledge
MMLU Pro
GLM 5.1
GLM 5.1 leads, 86.9% against 84.9%.
Same board, both measured
Web design
Design Arena: websites
GLM 5.1
GLM 5.1 leads, 1290 against 1175.
Same board, both measured
Long documents
Listed context window
GLM 5.1
GLM 5.1 takes 205K tokens against 164K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
DeepSeek V3.2
DeepSeek V3.2 $0.0007, GLM 5.1 $0.0036 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
GLM 5.1
DeepSeek V3.2 $0.99, GLM 5.1 $0.60.
Published attempt cost and pass rate

At a glance

Coding index: DeepSeek V3.2 30.0 from 3 benchmarks, GLM 5.1 49.2 from 3 benchmarks.
Widest category: Agentic work, 18.5 points apart on average over 3 shared benchmarks.
Widest single gap: Vending-Bench 2: DeepSeek V3.2 $1,034 against GLM 5.1 $5,634, 28.8 points apart on a 0 to 100 scale.
Output price: DeepSeek V3.2 $0.88 against GLM 5.1 $4.40 per million output tokens, 5.0x apart.
Context window: DeepSeek V3.2 takes 164K tokens, GLM 5.1 205K.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
Category	DeepSeek V3.2	GLM 5.1
Coding	0	10
Intelligence	0	2
Agentic work	0	3
Reasoning	0	6
Math	0	3
Knowledge	1	8
Web design	0	8
General	0	19
All shared boards	1	59

CodingLike-for-likeSix of these boards feed the Vibe Coding Index010

Coding benchmarks for DeepSeek V3.2 and GLM 5.1
Benchmark	DeepSeek V3.2	GLM 5.1	Leader
SWE-bench Verified	70.0%	Not measured	Only DeepSeek V3.2 measured
SWE-bench Multilingual	59.0%	Not measured	Only DeepSeek V3.2 measured
LMArena WebDev	1362	1508	GLM 5.1
AA Coding Index	44.2	55.8	GLM 5.1
LiveCodeBench (Vals AI)	80.7%	81.4%	GLM 5.1
SWE-bench Verified (Vals AI)	67.6%	76.4%	GLM 5.1
Vibe Code Bench (Vals AI)	5.1%	31.5%	GLM 5.1
Code migration (Vals AI)	Not measured	25.8%	Only GLM 5.1 measured
ProgramBench (Vals AI)	Not measured	0.0%	Only GLM 5.1 measured
Terminal-Bench 2.1 (Vals AI)	Not measured	56.9%	Only GLM 5.1 measured
LiveCodeBench (Artificial Analysis run)	86.2%	Not measured	Only DeepSeek V3.2 measured
SciCode	Not measured	44.8%	Only GLM 5.1 measured
Terminal-Bench Hard (Artificial Analysis run)	35.6%	43.2%	GLM 5.1
Terminal-Bench	39.6%	Not measured	Only DeepSeek V3.2 measured
WeirdML	Not measured	57.1%	Only GLM 5.1 measured
ALE-Bench	Not measured	887.1	Only GLM 5.1 measured
LMArena Text: coding	1453	1485	GLM 5.1
LMArena Text: software and IT	1449	1484	GLM 5.1
LMArena WebDev: React	1350	1505	GLM 5.1
LMArena WebDev: HTML	1371	1508	GLM 5.1
Neither model has a result on the other 19 boards here

IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02

Intelligence benchmarks for DeepSeek V3.2 and GLM 5.1
Benchmark	DeepSeek V3.2	GLM 5.1	Leader
AA Intelligence Index	21.5	26.1	GLM 5.1
Epoch Capabilities Index	146.1	149.8	GLM 5.1

Agentic workLike-for-likePublic boards, shown as published, kept out of the index03

Agentic work benchmarks for DeepSeek V3.2 and GLM 5.1
Benchmark	DeepSeek V3.2	GLM 5.1	Leader
AA Agentic Index	Not measured	23.9	Only GLM 5.1 measured
Design Arena: agents building web apps	Not measured	1175	Only GLM 5.1 measured
Design Arena: agents building full-stack apps	Not measured	1167	Only GLM 5.1 measured
Design Arena: agents building mobile apps	Not measured	1162	Only GLM 5.1 measured
Design Arena: agents building Android apps	Not measured	1180	Only GLM 5.1 measured
Design Arena: agents building Godot games	Not measured	1053	Only GLM 5.1 measured
Design Arena: agents building games	Not measured	1164	Only GLM 5.1 measured
Design Arena: HTML slides	Not measured	1170	Only GLM 5.1 measured
Design Arena: PowerPoint slides	Not measured	1188	Only GLM 5.1 measured
Design Arena: spreadsheets	Not measured	1315	Only GLM 5.1 measured
Finance Agent (Vals AI)	Not measured	44.8%	Only GLM 5.1 measured
Legal Research Bench (Vals AI)	Not measured	27.9%	Only GLM 5.1 measured
τ²-Bench Telecom	90.6%	97.7%	GLM 5.1
APEX-Agents	21.3%	40.9%	GLM 5.1
Vending-Bench 2	$1,034	$5,634	GLM 5.1
Neither model has a result on the other 26 boards here

ReasoningLike-for-likePublic boards, shown as published, kept out of the index06

Reasoning benchmarks for DeepSeek V3.2 and GLM 5.1
Benchmark	DeepSeek V3.2	GLM 5.1	Leader
GPQA Diamond	80.3%	84.5%	GLM 5.1
GPQA Diamond (Artificial Analysis run)	84.0%	86.8%	GLM 5.1
Humanity's Last Exam (Artificial Analysis run)	24.6%	30.1%	GLM 5.1
GPQA Diamond (Epoch AI run)	Not measured	89.9%	Only GLM 5.1 measured
ARC-AGI	57.0%	Not measured	Only DeepSeek V3.2 measured
ARC-AGI-2	4.0%	Not measured	Only DeepSeek V3.2 measured
SimpleBench	Not measured	55.1%	Only GLM 5.1 measured
Chess puzzles	Not measured	20.0%	Only GLM 5.1 measured
LMArena Text: hard prompts	1434	1473	GLM 5.1
LMArena Text: expert	1436	1478	GLM 5.1
LMArena Text: hard prompts in English	1446	1482	GLM 5.1
Neither model has a result on the other 11 boards here

MathLike-for-likePublic boards, shown as published, kept out of the index03

Math benchmarks for DeepSeek V3.2 and GLM 5.1
Benchmark	DeepSeek V3.2	GLM 5.1	Leader
AIME	84.6%	91.9%	GLM 5.1
MGSM (Vals AI)	86.0%	Not measured	Only DeepSeek V3.2 measured
AA Math Index	92.0	Not measured	Only DeepSeek V3.2 measured
AIME 2025 (Artificial Analysis run)	92.0%	Not measured	Only DeepSeek V3.2 measured
OTIS Mock AIME	Not measured	93.3%	Only GLM 5.1 measured
FrontierMath, tiers 1 to 3	Not measured	36.8%	Only GLM 5.1 measured
ProofBench	Not measured	22.2%	Only GLM 5.1 measured
LMArena Text: math	1435	1473	GLM 5.1
LMArena Text: mathematical work	1440	1470	GLM 5.1
Neither model has a result on the other 4 boards here

KnowledgeLike-for-likePublic boards, shown as published, kept out of the index18

Knowledge benchmarks for DeepSeek V3.2 and GLM 5.1
Benchmark	DeepSeek V3.2	GLM 5.1	Leader
MMLU Pro	84.9%	86.9%	GLM 5.1
LegalBench (Vals AI)	76.1%	84.4%	GLM 5.1
CaseLaw (Vals AI)	55.4%	51.5%	DeepSeek V3.2
CorpFin (Vals AI)	51.0%	64.5%	GLM 5.1
TaxEval (Vals AI)	68.2%	71.2%	GLM 5.1
MedQA (Vals AI)	93.9%	Not measured	Only DeepSeek V3.2 measured
MedCode (Vals AI)	Not measured	41.6%	Only GLM 5.1 measured
MMLU Pro (Artificial Analysis run)	86.2%	Not measured	Only DeepSeek V3.2 measured
SimpleQA Verified	Not measured	34.0%	Only GLM 5.1 measured
LMArena Text: science	1446	1478	GLM 5.1
LMArena Text: medicine and healthcare	1440	1474	GLM 5.1
LMArena Text: legal and government	1425	1465	GLM 5.1
LMArena Text: business and finance	1410	1452	GLM 5.1
Neither model has a result on the other 5 boards here

Web designLike-for-likePublic boards, shown as published, kept out of the index08

Web design benchmarks for DeepSeek V3.2 and GLM 5.1
Benchmark	DeepSeek V3.2	GLM 5.1	Leader
Design Arena: websites	1175	1290	GLM 5.1
Design Arena: UI components	1149	1292	GLM 5.1
Design Arena: game dev	1132	1265	GLM 5.1
Design Arena: data viz	1158	1366	GLM 5.1
Design Arena: 3D	1133	1336	GLM 5.1
Design Arena: all code categories	1164	1268	GLM 5.1
Design Arena: SVG	1032	1218	GLM 5.1
Design Arena: ASCII art	1083	1142	GLM 5.1
Neither model has a result on the other 1 board here

GeneralLike-for-likePublic boards, shown as published, kept out of the index019

General benchmarks for DeepSeek V3.2 and GLM 5.1
Benchmark	DeepSeek V3.2	GLM 5.1	Leader
Public Benefits Bench (Vals AI)	Not measured	61.8%	Only GLM 5.1 measured
MedScribe (Vals AI)	Not measured	72.3%	Only GLM 5.1 measured
IFBench	60.7%	76.3%	GLM 5.1
Long context reasoning (AA-LCR)	73.3%	73.7%	GLM 5.1
DTBench	85.6%	Not measured	Only DeepSeek V3.2 measured
LMCA	28.8%	Not measured	Only DeepSeek V3.2 measured
CL-bench	12.4%	Not measured	Only DeepSeek V3.2 measured
ExploitBench	Not measured	18.1%	Only GLM 5.1 measured
CL-bench Life	7.4%	Not measured	Only DeepSeek V3.2 measured
LMArena Text	1425	1463	GLM 5.1
LMArena Text: instruction following	1413	1453	GLM 5.1
LMArena Text: creative writing	1400	1453	GLM 5.1
LMArena Text: English	1439	1476	GLM 5.1
LMArena Text: non-English	1409	1448	GLM 5.1
LMArena Text: Chinese	1455	1515	GLM 5.1
LMArena Text: German	1413	1467	GLM 5.1
LMArena Text: Japanese	1374	1435	GLM 5.1
LMArena Text: Korean	1370	1424	GLM 5.1
LMArena Text: French	1429	1470	GLM 5.1
LMArena Text: Spanish	1424	1468	GLM 5.1
LMArena Text: Russian	1424	1454	GLM 5.1
LMArena Text: Polish	1419	1449	GLM 5.1
LMArena Text: multi-turn	1427	1473	GLM 5.1
LMArena Text: longer queries	1428	1467	GLM 5.1
LMArena Text: writing and language	1407	1454	GLM 5.1
LMArena Text: entertainment and media	1395	1442	GLM 5.1
Neither model has a result on the other 20 boards here

Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared

Image generation benchmarks for DeepSeek V3.2 and GLM 5.1
Benchmark	DeepSeek V3.2	GLM 5.1	Leader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

DeepSeek V3.2$0.0007Fits in one request

GLM 5.1$0.0036Fits in one request

DeepSeek V3.2 costs less

Repository review

50K input + 3K output tokens

DeepSeek V3.2$0.017Fits in one request

GLM 5.1$0.083Fits in one request

DeepSeek V3.2 costs less

Agent loop

200K input + 10K output tokens

DeepSeek V3.2$0.067Exceeds the context window

GLM 5.1$0.324Exceeds the context window

DeepSeek V3.2 costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of DeepSeek V3.2 and GLM 5.1
Specification	DeepSeek V3.2	Z.ai GLM 5.1
Maker	DeepSeek	Z.ai
Weights	Open weights	Open weights
Released	December 1, 2025	April 7, 2026
Context window	164K	205K
Input $/Mtok	$0.29	$1.40
Output $/Mtok	$0.88	$4.40
Coding index	30.0	49.2
Independent measurements behind it	3	3
Cost per solved task	$0.99	$0.60
Source	OpenRouter listing	OpenRouter listing

Questions

Is DeepSeek V3.2 better than GLM 5.1 for coding?

Pick GLM 5.1 for coding: it scores 49.2 on the Vibe Coding Index against 30.0 for DeepSeek V3.2. Pick DeepSeek V3.2 for output tokens 5.0x cheaper ($0.88 vs $4.40 per million) and work measured by CaseLaw (Vals AI), where it leads.

Which is cheaper, DeepSeek V3.2 or GLM 5.1?

DeepSeek V3.2 costs $0.29 per million input tokens and $0.88 per million output tokens; GLM 5.1 costs $1.40 and $4.40. Per solved SWE-bench Verified task, DeepSeek V3.2 costs $0.99 and GLM 5.1 $0.60.

Which has the bigger context window, DeepSeek V3.2 or GLM 5.1?

GLM 5.1, with 205K tokens against 164K.

How many benchmarks is the DeepSeek V3.2 vs GLM 5.1 comparison based on?

60 benchmarks have a result for both models as of October 21, 2026. DeepSeek V3.2's coding index rests on 3 independent benchmarks and GLM 5.1's on 3. Where only one model has a result, the page says so instead of naming a winner.

Compare next

Embed this comparison on your site

Free under CC BY 4.0, while the source link stays visible. Preview the frame · Terms and other embeds

<iframe src="https://llmbenchmark.ai/embed/compare/deepseek-v3-2-vs-glm-5-1" title="DeepSeek V3.2 vs GLM 5.1, LLMBenchmark.ai" width="100%" height="360" style="border:0;max-width:100%" loading="lazy"></iframe>
<p style="font:13px/1.4 sans-serif;margin:6px 0 0">Source: <a href="https://llmbenchmark.ai/compare/deepseek-v3-2-vs-glm-5-1">LLMBenchmark.ai, DeepSeek V3.2 vs GLM 5.1</a></p>

DeepSeek V3.2 or GLM 5.1: which to pick

Which one for your work

Coding

Intelligence

Agentic work

Reasoning

Math

Knowledge

Web design

Long documents

Everyday cost

Cost per solved task

At a glance

Benchmark by benchmark

What each workload costs

Chat turn

Repository review

Agent loop

Specification differences

Questions

Compare next