Enter opens the highlighted result. Esc closes.

Gemma 3 4B vs GLM 4.6

GLM 4.6 leads Gemma 3 4B on the Artificial Analysis Intelligence Index, 18.5 to 4.8, and Gemma 3 4B costs 29.3x less per output token ($0.075 vs $2.20 per million). Updated October 21, 2026; 36 benchmarks have a result for both.

Replace Gemma 3 4B Replace GLM 4.6
Gemma 3 4BGoogle
Coding
No data
Overall
22.6
Output $/M
$0.075
0
GLM 4.6Z.ai
Coding
8.8
Overall
33.4
Output $/M
$2.20
36
vs boards won of 36 shared

Gemma 3 4B or GLM 4.6: which to pick

Only GLM 4.6 has a coding score, 8.8 on the Vibe Coding Index; Gemma 3 4B has none, so the two cannot be ranked on coding.

Pick Gemma 3 4B for output tokens 29.3x cheaper ($0.075 vs $2.20 per million).

Model AGemma 3 4BGoogleNot benchmarkedNo coding benchmarkNo cost per solved taskModel BGLM 4.6Z.ai8.8/1002 independent benchmarksNo cost per solved task
Shared results
36
Gemma 3 4B only
6
GLM 4.6 only
33
Like-for-like categories
7 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.

At a glance

Coding index
Gemma 3 4B not benchmarked, GLM 4.6 8.8 from 2 benchmarks.
Widest category
Agentic work, 62.3 points apart on average over 2 shared benchmarks.
Widest single gap
AIME 2025 (Artificial Analysis run): Gemma 3 4B 12.7% against GLM 4.6 86.0%, 73.3 points apart on a 0 to 100 scale.
Output price
Gemma 3 4B $0.075 against GLM 4.6 $2.20 per million output tokens, 29.3x apart.
Context window
Gemma 3 4B takes 131K tokens, GLM 4.6 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
CategoryGemma 3 4BBarsGLM 4.6Ties
Coding05
Intelligence01
Agentic work02
Reasoning06
Math04
Knowledge05
General013
All shared boards036
CodingLike-for-likeSix of these boards feed the Vibe Coding Index05
Coding benchmarks for Gemma 3 4B and GLM 4.6
BenchmarkGemma 3 4BGLM 4.6Leader
SWE-bench VerifiedNot measured68.2%Only GLM 4.6 measured
LMArena WebDevNot measured1341Only GLM 4.6 measured
AA Coding Index2.745.8GLM 4.6
LiveCodeBench (Vals AI)Not measured81.0%Only GLM 4.6 measured
Vibe Code Bench (Vals AI)Not measured3.1%Only GLM 4.6 measured
LiveCodeBench (Artificial Analysis run)11.2%69.5%GLM 4.6
Terminal-Bench Hard (Artificial Analysis run)0.8%28.8%GLM 4.6
Terminal-BenchNot measured24.5%Only GLM 4.6 measured
ALE-BenchNot measured340.8Only GLM 4.6 measured
LMArena Text: coding12301449GLM 4.6
LMArena Text: software and IT12691455GLM 4.6
LMArena WebDev: HTMLNot measured1350Only GLM 4.6 measured
SciCode0.0%Not measuredOnly Gemma 3 4B measured
Neither model has a result on the other 26 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
Intelligence benchmarks for Gemma 3 4B and GLM 4.6
BenchmarkGemma 3 4BGLM 4.6Leader
AA Intelligence Index4.818.5GLM 4.6
Epoch Capabilities Index115.9Not measuredOnly Gemma 3 4B measured
Agentic workLike-for-likePublic boards, shown as published, kept out of the index02
Agentic work benchmarks for Gemma 3 4B and GLM 4.6
BenchmarkGemma 3 4BGLM 4.6Leader
Design Arena: agents building full-stack appsNot measured1029Only GLM 4.6 measured
Design Arena: agents building mobile appsNot measured1096Only GLM 4.6 measured
Design Arena: agents building Android appsNot measured1084Only GLM 4.6 measured
Design Arena: agents building Godot gamesNot measured1180Only GLM 4.6 measured
τ²-Bench Telecom5.0%76.9%GLM 4.6
Berkeley Function Calling Leaderboard (BFCL)19.6%72.4%GLM 4.6
BrowseComp-PlusNot measured48.2%Only GLM 4.6 measured
Neither model has a result on the other 34 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index06
Reasoning benchmarks for Gemma 3 4B and GLM 4.6
BenchmarkGemma 3 4BGLM 4.6Leader
GPQA DiamondNot measured74.5%Only GLM 4.6 measured
GPQA Diamond (Artificial Analysis run)29.1%78.0%GLM 4.6
Humanity's Last Exam (Artificial Analysis run)5.3%14.5%GLM 4.6
GPQA Diamond (Epoch AI run)23.2%Not measuredOnly Gemma 3 4B measured
LMArena Text: hard prompts12531440GLM 4.6
LMArena Text: expert12231432GLM 4.6
LMArena Text: hard prompts in English12531445GLM 4.6
Kagi LLM Benchmark25.2%45.7%GLM 4.6
Neither model has a result on the other 14 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index04
Math benchmarks for Gemma 3 4B and GLM 4.6
BenchmarkGemma 3 4BGLM 4.6Leader
AIMENot measured92.7%Only GLM 4.6 measured
MGSM (Vals AI)Not measured89.8%Only GLM 4.6 measured
AA Math Index12.786.0GLM 4.6
AIME 2025 (Artificial Analysis run)12.7%86.0%GLM 4.6
OTIS Mock AIME7.5%Not measuredOnly Gemma 3 4B measured
LMArena Text: math12391432GLM 4.6
LMArena Text: mathematical work12561447GLM 4.6
Neither model has a result on the other 6 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index05
Knowledge benchmarks for Gemma 3 4B and GLM 4.6
BenchmarkGemma 3 4BGLM 4.6Leader
MMLU ProNot measured82.2%Only GLM 4.6 measured
LegalBench (Vals AI)Not measured79.6%Only GLM 4.6 measured
CorpFin (Vals AI)Not measured56.8%Only GLM 4.6 measured
TaxEval (Vals AI)Not measured66.2%Only GLM 4.6 measured
MedQA (Vals AI)Not measured92.2%Only GLM 4.6 measured
MMLU Pro (Artificial Analysis run)41.7%82.9%GLM 4.6
LMArena Text: science13061456GLM 4.6
LMArena Text: medicine and healthcare13421455GLM 4.6
LMArena Text: legal and government13271448GLM 4.6
LMArena Text: business and finance13111416GLM 4.6
Neither model has a result on the other 8 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index7 boards, none shared
Web design benchmarks for Gemma 3 4B and GLM 4.6
BenchmarkGemma 3 4BGLM 4.6Leader
Design Arena: websitesNot measured1174Only GLM 4.6 measured
Design Arena: UI componentsNot measured1148Only GLM 4.6 measured
Design Arena: game devNot measured1151Only GLM 4.6 measured
Design Arena: data vizNot measured1161Only GLM 4.6 measured
Design Arena: 3DNot measured1129Only GLM 4.6 measured
Design Arena: all code categoriesNot measured1166Only GLM 4.6 measured
Design Arena: SVGNot measured1109Only GLM 4.6 measured
Neither model has a result on the other 2 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index013
General benchmarks for Gemma 3 4B and GLM 4.6
BenchmarkGemma 3 4BGLM 4.6Leader
IFBench28.3%43.4%GLM 4.6
Long context reasoning (AA-LCR)6.7%54.0%GLM 4.6
DTBench50.9%Not measuredOnly Gemma 3 4B measured
LMCA2.8%Not measuredOnly Gemma 3 4B measured
LMArena Text12911440GLM 4.6
LMArena Text: instruction following12391410GLM 4.6
LMArena Text: creative writing12701412GLM 4.6
LMArena Text: English13001449GLM 4.6
LMArena Text: non-English12731426GLM 4.6
LMArena Text: ChineseNot measured1499Only GLM 4.6 measured
LMArena Text: German12811447GLM 4.6
LMArena Text: JapaneseNot measured1393Only GLM 4.6 measured
LMArena Text: KoreanNot measured1400Only GLM 4.6 measured
LMArena Text: FrenchNot measured1459Only GLM 4.6 measured
LMArena Text: SpanishNot measured1436Only GLM 4.6 measured
LMArena Text: Russian12941418GLM 4.6
LMArena Text: PolishNot measured1447Only GLM 4.6 measured
LMArena Text: multi-turn12551427GLM 4.6
LMArena Text: longer queries12731422GLM 4.6
LMArena Text: writing and language12881415GLM 4.6
LMArena Text: entertainment and media12331409GLM 4.6
Neither model has a result on the other 25 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Gemma 3 4B and GLM 4.6
BenchmarkGemma 3 4BGLM 4.6Leader
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

Gemma 3 4B$0.0001Fits in one request
GLM 4.6$0.0017Fits in one request

Gemma 3 4B costs less

Repository review

50K input + 3K output tokens

Gemma 3 4B$0.0025Fits in one request
GLM 4.6$0.037Fits in one request

Gemma 3 4B costs less

Agent loop

200K input + 10K output tokens

Gemma 3 4B$0.0097Exceeds the context window
GLM 4.6$0.142Exceeds the context window

Gemma 3 4B 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 Gemma 3 4B and GLM 4.6
SpecificationGoogle Gemma 3 4BZ.ai GLM 4.6
MakerGoogleZ.ai
WeightsOpen weightsOpen weights
ReleasedMarch 13, 2025September 30, 2025
Context window131K205K
Input $/Mtok$0.045$0.60
Output $/Mtok$0.075$2.20
Coding indexNot benchmarked8.8
Independent measurements behind it02
Cost per solved taskNot computableNot computable
SourceOpenRouter listingOpenRouter listing

Questions

Is Gemma 3 4B better than GLM 4.6 for coding?

Only GLM 4.6 has a coding score, 8.8 on the Vibe Coding Index; Gemma 3 4B has none, so the two cannot be ranked on coding. Pick Gemma 3 4B for output tokens 29.3x cheaper ($0.075 vs $2.20 per million).

Which is cheaper, Gemma 3 4B or GLM 4.6?

Gemma 3 4B costs $0.045 per million input tokens and $0.075 per million output tokens; GLM 4.6 costs $0.60 and $2.20. Cost per solved task is known for neither of them.

Which has the bigger context window, Gemma 3 4B or GLM 4.6?

GLM 4.6, with 205K tokens against 131K.

How many benchmarks is the Gemma 3 4B vs GLM 4.6 comparison based on?

36 benchmarks have a result for both models as of October 21, 2026. Gemma 3 4B's coding index rests on 0 independent benchmarks and GLM 4.6's on 2. Where only one model has a result, the page says so instead of naming a winner.

Compare next