Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models741Deals30d271Policy30d67, 15 pendingSources1,014

Gemma 4 31B vs Qwen3.5 35B A3B

Google · released

Wins 5 of 8 areas

Coding · Agents · Reasoning · Facts · Following instructions

Alibaba · released

Wins 3 of 8 areas

Images and charts · Math · Long documents

Gemma 4 31B is the stronger all-rounder.Qwen3.5 35B A3B is better at images and charts.

Scores updated · 65 tests both models report · How we compare

Where each one wins

Tests won in each of the eight areas we test. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.

Long documents and following instructions rest on a single test each.

What it costs

Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.

Gemma 4 31B$0.17 to read · $0.40 to write$0.57Price from Artificial Analysis
Qwen3.5 35B A3B$0.25 to read · $2.00 to write$2.25Price from Alibaba's own price page

Gemma 4 31B costs 75% less for the same work.

The biggest differences

The three tests each model wins by the widest margin. Scores are out of 100.

Where Gemma 4 31B pulls ahead

  • Finds the root cause of IT system incidentsAA IT-Bench SRE+15.8points aheadGemma 4 31B37.3Qwen3.5 35B A3B21.5
  • Hard command-line tasks in a real terminalTerminal-Bench Hard+9.9points aheadGemma 4 31B36.4Qwen3.5 35B A3B26.5
  • Customer-service tasks in a simulated bankτ-Bench V3 · Banking+9.9points aheadGemma 4 31B14.8Qwen3.5 35B A3B4.9

Where Qwen3.5 35B A3B pulls ahead

  • Fixes real GitHub issues in Python projectsSWE-bench Verified+17.2points aheadQwen3.5 35B A3B69.2Gemma 4 31B52
  • Reasoning about charts from research papersCharXiv (RQ)+9.6points aheadQwen3.5 35B A3B77.5Gemma 4 31B67.9
  • Long, multi-file coding tasks in real codebasesSWE-bench Pro+8.9points aheadQwen3.5 35B A3B44.6Gemma 4 31B35.7

Every test, side by side

All 65 tests both models report. The winning score is in its model's colour; marks a score checked independently.

CodingGemma 4 31B

Full coding ranking

AgentsGemma 4 31B

Full agents ranking

ReasoningGemma 4 31B

Full reasoning ranking

FactsGemma 4 31B

Full facts ranking

Images and chartsQwen3.5 35B A3B

Full images and charts ranking

MathQwen3.5 35B A3B

Full math ranking

Long documentsQwen3.5 35B A3B
  • AA-LCRQwen3.5 35B A3B by 2.369.772+2.3

Full long documents ranking

Following instructionsGemma 4 31B
  • IFBenchGemma 4 31B by 3.175.672.5+3.1

Full following instructions ranking

Other results38 tests, not counted

Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.

Questions people ask

Which is better, Gemma 4 31B or Qwen3.5 35B A3B?

Gemma 4 31B wins five of the eight areas we test: coding, agents, reasoning, facts and following instructions. Qwen3.5 35B A3B wins images and charts, math and long documents.

Which is better for coding?

Gemma 4 31B. It wins 5 of the 8 coding tests both models report; Qwen3.5 35B A3B wins 3.

Which is cheaper?

Gemma 4 31B costs $0.17 per million input tokens and $0.40 per million output tokens; Qwen3.5 35B A3B costs $0.25 and $2.00. That makes Gemma 4 31B about 75% cheaper for the same work.

How do you compare the two?

We use the 65 benchmark tests both models have published scores on. The verdict counts the 27 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 38 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.

Compare Gemma 4 31B with

Google

Across the field

Compare Qwen3.5 35B A3B with

Alibaba

Across the field

All comparisons