Enter opens the highlighted result. Esc closes.

Gemma 3 4B vs Qwen3.5-122B-A10B

Qwen3.5-122B-A10B leads Gemma 3 4B on the Artificial Analysis Intelligence Index, 17.7 to 4.8, and Gemma 3 4B costs 27.7x less per output token ($0.075 vs $2.08 per million). Updated October 22, 2026; 32 benchmarks have a result for both.

Replace Gemma 3 4B Replace Qwen3.5-122B-A10B
Gemma 3 4BGoogle
Coding
No data
Overall
22.6
Output $/M
$0.075
0
Qwen3.5-122B-A10BQwen
Coding
No data
Overall
No data
Output $/M
$2.08
32
vs boards won of 32 shared

Gemma 3 4B or Qwen3.5-122B-A10B: which to pick

Neither model has a coding score yet, so the two cannot be ranked on coding.

Pick Gemma 3 4B for output tokens 27.7x cheaper ($0.075 vs $2.08 per million) and work measured by AA Coding Index or Terminal-Bench Hard (Artificial Analysis run), where it leads.

Model AGemma 3 4BGoogleNot benchmarkedNo coding benchmarkNo cost per solved taskModel BQwen3.5-122B-A10BQwenNot benchmarked1 independent benchmarkNo cost per solved task
Shared results
32
Gemma 3 4B only
10
Qwen3.5-122B-A10B only
19
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, Qwen3.5-122B-A10B not benchmarked.
Widest category
Agentic work, 88.6 points apart on average over 1 shared benchmark.
Widest single gap
τ²-Bench Telecom: Gemma 3 4B 5.0% against Qwen3.5-122B-A10B 93.6%, 88.6 points apart on a 0 to 100 scale.
Output price
Gemma 3 4B $0.075 against Qwen3.5-122B-A10B $2.08 per million output tokens, 27.7x apart.
Context window
Gemma 3 4B takes 131K tokens, Qwen3.5-122B-A10B 262K.

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 4BBarsQwen3.5-122B-A10BTies
Coding04
Intelligence01
Agentic work01
Reasoning05
Math02
Knowledge04
General015
All shared boards032
CodingLike-for-likeSix of these boards feed the Vibe Coding Index04
Coding benchmarks for Gemma 3 4B and Qwen3.5-122B-A10B
BenchmarkGemma 3 4BQwen3.5-122B-A10BLeader
LMArena WebDevNot measured1361Only Qwen3.5-122B-A10B measured
AA Coding Index2.743.3Qwen3.5-122B-A10B
LiveCodeBench (Artificial Analysis run)11.2%Not measuredOnly Gemma 3 4B measured
SciCodeNot measured39.7%Only Qwen3.5-122B-A10B measured
Terminal-Bench Hard (Artificial Analysis run)0.8%31.1%Qwen3.5-122B-A10B
LMArena Text: coding12301436Qwen3.5-122B-A10B
LMArena Text: software and IT12691436Qwen3.5-122B-A10B
LMArena WebDev: ReactNot measured1353Only Qwen3.5-122B-A10B measured
LMArena WebDev: HTMLNot measured1351Only Qwen3.5-122B-A10B measured
SciCode0.0%Not measuredOnly Gemma 3 4B measured
Neither model has a result on the other 33 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
Intelligence benchmarks for Gemma 3 4B and Qwen3.5-122B-A10B
BenchmarkGemma 3 4BQwen3.5-122B-A10BLeader
AA Intelligence Index4.817.7Qwen3.5-122B-A10B
Epoch Capabilities Index115.9Not measuredOnly Gemma 3 4B measured
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
Agentic work benchmarks for Gemma 3 4B and Qwen3.5-122B-A10B
BenchmarkGemma 3 4BQwen3.5-122B-A10BLeader
τ²-Bench Telecom5.0%93.6%Qwen3.5-122B-A10B
Berkeley Function Calling Leaderboard (BFCL)19.6%Not measuredOnly Gemma 3 4B measured
Neither model has a result on the other 40 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index05
Reasoning benchmarks for Gemma 3 4B and Qwen3.5-122B-A10B
BenchmarkGemma 3 4BQwen3.5-122B-A10BLeader
GPQA Diamond (Artificial Analysis run)29.1%85.7%Qwen3.5-122B-A10B
Humanity's Last Exam (Artificial Analysis run)5.3%25.2%Qwen3.5-122B-A10B
GPQA Diamond (Epoch AI run)23.2%Not measuredOnly Gemma 3 4B measured
Mystery game puzzlesNot measured17.0%Only Qwen3.5-122B-A10B measured
LMArena Text: hard prompts12531421Qwen3.5-122B-A10B
LMArena Text: expert12231431Qwen3.5-122B-A10B
LMArena Text: hard prompts in English12531431Qwen3.5-122B-A10B
LMArena Vision: diagramsNot measured1249Only Qwen3.5-122B-A10B measured
LMArena Vision: homeworkNot measured1261Only Qwen3.5-122B-A10B measured
Kagi LLM Benchmark25.2%Not measuredOnly Gemma 3 4B measured
Neither model has a result on the other 13 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index02
Math benchmarks for Gemma 3 4B and Qwen3.5-122B-A10B
BenchmarkGemma 3 4BQwen3.5-122B-A10BLeader
AA Math Index12.7Not measuredOnly Gemma 3 4B measured
AIME 2025 (Artificial Analysis run)12.7%Not measuredOnly Gemma 3 4B measured
OTIS Mock AIME7.5%Not measuredOnly Gemma 3 4B measured
LMArena Text: math12391431Qwen3.5-122B-A10B
LMArena Text: mathematical work12561431Qwen3.5-122B-A10B
Neither model has a result on the other 9 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index04
Knowledge benchmarks for Gemma 3 4B and Qwen3.5-122B-A10B
BenchmarkGemma 3 4BQwen3.5-122B-A10BLeader
MMLU Pro (Artificial Analysis run)41.7%Not measuredOnly Gemma 3 4B measured
LMArena Text: science13061432Qwen3.5-122B-A10B
LMArena Text: medicine and healthcare13421423Qwen3.5-122B-A10B
LMArena Text: legal and government13271422Qwen3.5-122B-A10B
LMArena Text: business and finance13111414Qwen3.5-122B-A10B
Neither model has a result on the other 15 boards here
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Web design benchmarks for Gemma 3 4B and Qwen3.5-122B-A10B
BenchmarkGemma 3 4BQwen3.5-122B-A10BLeader
Neither model has a result on 9 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index015
General benchmarks for Gemma 3 4B and Qwen3.5-122B-A10B
BenchmarkGemma 3 4BQwen3.5-122B-A10BLeader
IFBench28.3%75.7%Qwen3.5-122B-A10B
Long context reasoning (AA-LCR)6.7%76.3%Qwen3.5-122B-A10B
DTBench50.9%84.3%Qwen3.5-122B-A10B
LMCA2.8%32.2%Qwen3.5-122B-A10B
LMArena Text12911417Qwen3.5-122B-A10B
LMArena Text: instruction following12391399Qwen3.5-122B-A10B
LMArena Text: creative writing12701368Qwen3.5-122B-A10B
LMArena VisionNot measured1244Only Qwen3.5-122B-A10B measured
LMArena Text: English13001428Qwen3.5-122B-A10B
LMArena Text: non-English12731401Qwen3.5-122B-A10B
LMArena Text: ChineseNot measured1463Only Qwen3.5-122B-A10B measured
LMArena Text: German12811425Qwen3.5-122B-A10B
LMArena Text: JapaneseNot measured1367Only Qwen3.5-122B-A10B measured
LMArena Text: KoreanNot measured1353Only Qwen3.5-122B-A10B measured
LMArena Text: FrenchNot measured1443Only Qwen3.5-122B-A10B measured
LMArena Text: SpanishNot measured1425Only Qwen3.5-122B-A10B measured
LMArena Text: Russian12941400Qwen3.5-122B-A10B
LMArena Text: PolishNot measured1398Only Qwen3.5-122B-A10B measured
LMArena Text: multi-turn12551416Qwen3.5-122B-A10B
LMArena Text: longer queries12731410Qwen3.5-122B-A10B
LMArena Text: writing and language12881387Qwen3.5-122B-A10B
LMArena Text: entertainment and media12331366Qwen3.5-122B-A10B
LMArena Vision: EnglishNot measured1249Only Qwen3.5-122B-A10B measured
LMArena Vision: ChineseNot measured1301Only Qwen3.5-122B-A10B measured
LMArena Vision: reading textNot measured1251Only Qwen3.5-122B-A10B measured
LMArena Vision: humorNot measured1243Only Qwen3.5-122B-A10B measured
LMArena Vision: creative writingNot measured1220Only Qwen3.5-122B-A10B measured
Neither model has a result on the other 21 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 Qwen3.5-122B-A10B
BenchmarkGemma 3 4BQwen3.5-122B-A10BLeader
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
Qwen3.5-122B-A10B$0.0013Fits in one request

Gemma 3 4B costs less

Repository review

50K input + 3K output tokens

Gemma 3 4B$0.0025Fits in one request
Qwen3.5-122B-A10B$0.019Fits in one request

Gemma 3 4B costs less

Agent loop

200K input + 10K output tokens

Gemma 3 4B$0.0097Exceeds the context window
Qwen3.5-122B-A10B$0.073Fits in one request

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 Qwen3.5-122B-A10B
SpecificationGoogle Gemma 3 4BQwen3.5-122B-A10B
MakerGoogleQwen
WeightsOpen weightsOpen weights
ReleasedMarch 13, 2025February 25, 2026
Context window131K262K
Input $/Mtok$0.045$0.26
Output $/Mtok$0.075$2.08
Coding indexNot benchmarkedNot benchmarked
Independent measurements behind it01
Cost per solved taskNot computableNot computable
SourceOpenRouter listingOpenRouter listing

Questions

Is Gemma 3 4B better than Qwen3.5-122B-A10B for coding?

Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Gemma 3 4B for output tokens 27.7x cheaper ($0.075 vs $2.08 per million) and work measured by AA Coding Index or Terminal-Bench Hard (Artificial Analysis run), where it leads.

Which is cheaper, Gemma 3 4B or Qwen3.5-122B-A10B?

Gemma 3 4B costs $0.045 per million input tokens and $0.075 per million output tokens; Qwen3.5-122B-A10B costs $0.26 and $2.08. Cost per solved task is known for neither of them.

Which has the bigger context window, Gemma 3 4B or Qwen3.5-122B-A10B?

Qwen3.5-122B-A10B, with 262K tokens against 131K.

How many benchmarks is the Gemma 3 4B vs Qwen3.5-122B-A10B comparison based on?

32 benchmarks have a result for both models as of October 22, 2026. Gemma 3 4B's coding index rests on 0 independent benchmarks and Qwen3.5-122B-A10B's on 1. Where only one model has a result, the page says so instead of naming a winner.

Compare next