Gemma 4 31B vs Qwen3.5 397B

Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.

Agentic
Coding
Multimodal & Grounded
Reasoning
Knowledge
Instruction Following
Multilingual
Mathematics

Gemma 4 31B· Qwen3.5 397B

Quick Verdict

Pick Gemma 4 31B if you want the stronger benchmark profile. Qwen3.5 397B only becomes the better choice if knowledge is the priority or you would rather avoid the extra latency and token burn of a reasoning model.

Gemma 4 31B is clearly ahead on the aggregate, 73 to 68. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

Gemma 4 31B's sharpest advantage is in coding, where it averages 80 against 52.2. The single biggest benchmark swing on the page is LiveCodeBench, 80% to 39%. Qwen3.5 397B does hit back in knowledge, so the answer changes if that is the part of the workload you care about most.

Gemma 4 31B is the reasoning model in the pair, while Qwen3.5 397B is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use. Gemma 4 31B gives you the larger context window at 256K, compared with 128K for Qwen3.5 397B.

Operational tradeoffs

PriceFree*Free*
SpeedN/A96 t/s
TTFTN/A2.44s
Context256K128K

Decision framing

BenchLM keeps the benchmark table and the operator tradeoffs on the same page so a better score does not hide a materially slower, pricier, or smaller-context model.

Runtime metrics show N/A when BenchLM does not have a sourced snapshot for that exact model. The scoring rules and freshness policy are documented on the methodology page.

BenchmarkGemma 4 31BQwen3.5 397B
Agentic
Terminal-Bench 2.052.5%
BrowseComp62%
OSWorld-Verified62.2%
Claw-Eval48.1%
QwenClawBench51.8%
QwenWebBench1162
TAU3-Bench68.4%
DeepPlanning37.6%
Toolathlon36.3%
MCP Atlas46.1%
MCP-Tasks74.2%
WideResearch74.0%
CodingGemma 4 31B wins
LiveCodeBench80%39%
HumanEval75%
SWE-bench Verified76.2%
LiveCodeBench v683.6%
SWE-bench Pro50.9%
SWE Multilingual69.3%
NL2Repo32.2%
Multimodal & GroundedGemma 4 31B wins
MMMU-Pro76.9%79%
OfficeQA Pro68%
RealWorldQA83.9%
Video-MME (w/o subtitle)84.2%
MathVision88.6%
We-Math87.9%
DynaMath86.3%
MStar83.8%
SimpleVQA67.1%
ChatCVQA80.8%
AI2D_TEST93.9%
CountBench97.2%
RefCOCO (avg)92.3%
ODINW1347.0%
MLVU (M-Avg)86.7%
ScreenSpot Pro65.6%
ReasoningQwen3.5 397B wins
BBH74.4%82%
MRCRv266.4%71%
MuSR78%
LongBench v263.2%
AI-Needle68.7%
KnowledgeQwen3.5 397B wins
GPQA84.3%88.4%
MMLU-Pro85.2%87.8%
HLE26.5%28.7%
HLE w/o tools19.5%
MMLU83%
SuperGPQA70.4%
MMLU-Redux94.9%
C-Eval93%
FrontierScience71%
SimpleQA80%
Instruction Following
IFEval92.6%
IFBench76.5%
Multilingual
MGSM82%
MMLU-ProX84.7%
NOVA-6359.1%
INCLUDE85.6%
PolyMath73.3%
VWT2k-lite78.9%
MAXIFE88.2%
Mathematics
AIME 202383%
AIME 202485%
AIME 202584%
AIME2693.3%
HMMT Feb 202379%
HMMT Feb 202481%
HMMT Feb 202580%
HMMT Feb 202594.8%
HMMT Nov 202592.7%
HMMT Feb 202687.9%
MMAnswerBench80.9%
BRUMO 202582%
MATH-50081%
Frequently Asked Questions (5)

Which is better, Gemma 4 31B or Qwen3.5 397B?

Gemma 4 31B is ahead overall, 73 to 68. The biggest single separator in this matchup is LiveCodeBench, where the scores are 80% and 39%.

Which is better for knowledge tasks, Gemma 4 31B or Qwen3.5 397B?

Qwen3.5 397B has the edge for knowledge tasks in this comparison, averaging 68.2 versus 61.3. Inside this category, GPQA is the benchmark that creates the most daylight between them.

Which is better for coding, Gemma 4 31B or Qwen3.5 397B?

Gemma 4 31B has the edge for coding in this comparison, averaging 80 versus 52.2. Inside this category, LiveCodeBench is the benchmark that creates the most daylight between them.

Which is better for reasoning, Gemma 4 31B or Qwen3.5 397B?

Qwen3.5 397B has the edge for reasoning in this comparison, averaging 69.7 versus 66.4. Inside this category, BBH is the benchmark that creates the most daylight between them.

Which is better for multimodal and grounded tasks, Gemma 4 31B or Qwen3.5 397B?

Gemma 4 31B has the edge for multimodal and grounded tasks in this comparison, averaging 76.9 versus 74.1. Inside this category, MMMU-Pro is the benchmark that creates the most daylight between them.

Last updated: April 2, 2026

Weekly LLM Benchmark Digest

Get notified when new models drop, benchmark scores change, or the leaderboard shifts. One email per week.

Free. No spam. Unsubscribe anytime. We only store derived location metadata for consent routing.