Skip to content
VECTOR WIREAI INTELLIGENCE
UTC

Gemma 4 31B vs Nemotron 3.5 Super VL

Google · released

Wins 0 of 5 areas

—

NVIDIA · released

Wins 5 of 5 areas

Coding · Agents · Reasoning · Facts · Long documents

Nemotron 3.5 Super VL is the stronger all-rounder.Gemma 4 31B is cheaper.

Scores updated · 36 tests both models report · How we compare

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$0.15 to read · $0.40 to write$0.55Price from Artificial Analysis
Nemotron 3.5 Super VL$0.15 to read · $0.50 to write$0.65Price from Artificial Analysis

Gemma 4 31B costs 15% less for the same work.

The biggest differences

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

Where Nemotron 3.5 Super VL pulls ahead

  • Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+49.2points aheadNemotron 3.5 Super VL64.2Gemma 4 31B15
  • Finds hard-to-locate facts by browsing the webBrowseComp+41.9points aheadNemotron 3.5 Super VL78.9Gemma 4 31B37
  • Real work tasks from 44 professionsGDPVal+27.1points aheadNemotron 3.5 Super VL33.1Gemma 4 31B6

Where Gemma 4 31B pulls ahead

No clear win on a test scored out of 100.

Every test, side by side

All 36 tests both models report. The winning score is in its model's colour; marks a score checked independently.

CodingNemotron 3.5 Super VL

Full coding ranking

AgentsNemotron 3.5 Super VL
  • BrowseCompNemotron 3.5 Super VL by 41.93778.9+41.9
  • GDPValNemotron 3.5 Super VL by 27.1633.1+27.1
  • MCP AtlasNemotron 3.5 Super VL by 13.554.267.7+13.5
  • Terminal-Bench 4.0tie00tie

Full agents ranking

ReasoningNemotron 3.5 Super VL

Full reasoning ranking

FactsNemotron 3.5 Super VL

Full facts ranking

Long documentsNemotron 3.5 Super VL

Full long documents ranking

Other results22 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.

  • GDPval-AA 2.1Nemotron 3.5 Super VL by 535 rating points6271162+535 rating
  • AA-OmniscienceNemotron 3.5 Super VL by 43.7-47.9-4.2+43.7
  • ScreenSpot Pro originalNemotron 3.5 Super VL by 37.65.943.5+37.6
  • SWEBench Pro v2 (Hard)Nemotron 3.5 Super VL by 29.914.244.1+29.9
  • JobBenchNemotron 3.5 Super VL by 28.38.236.5+28.3
  • Misbehavior (Internal) ↓Nemotron 3.5 Super VL by 25.227.652.8+25.2
  • Covert misbehavior (Internal) ↓Nemotron 3.5 Super VL by 19.124.343.4+19.1
  • DeepSearchQA (F1)Nemotron 3.5 Super VL by 1661.777.7+16
  • AA IntelligenceNemotron 3.5 Super VL by 11.814.726.5+11.8
  • AutomationBench-AA v1.0.6Nemotron 3.5 Super VL by 9.94.914.8+9.9
  • CritPt v2 HardNemotron 3.5 Super VL by 9.5413.5+9.5
  • BVBNemotron 3.5 Super VL by 5.14752.1+5.1
  • OmniDocBench v1.6Gemma 4 31B by 4.682.577.9+4.6
  • Chartography originalGemma 4 31B by 2.3129.7+2.3
  • GDP (Surge AI)Nemotron 3.5 Super VL by 268+2
  • ZeroBench (main) originalNemotron 3.5 Super VL by 2911+2
  • HLE-DiamondNemotron 3.5 Super VL by 1.67.28.8+1.6
  • ComplexConstraints v1.1Gemma 4 31B by 1.531.630.1+1.5
  • StrongREJECT (none)Nemotron 3.5 Super VL by 1.398.199.4+1.3
  • MMMU Pro Standard 10Gemma 4 31B by 1.173.472.3+1.1
  • Terminal-Bench 4.0tie00tie
  • Terminal-Bench-Science 0.1tie00tie

Questions people ask

Which is better, Gemma 4 31B or Nemotron 3.5 Super VL?

Nemotron 3.5 Super VL wins all five areas where both have results: coding, agents, reasoning, facts and long documents. Gemma 4 31B wins none, but costs 15% less.

Which is better for coding?

Nemotron 3.5 Super VL. It wins 4 of the 4 coding tests both models report; Gemma 4 31B wins none.

Which is cheaper?

Gemma 4 31B costs $0.15 per million input tokens and $0.40 per million output tokens; Nemotron 3.5 Super VL costs $0.15 and $0.50. That makes Gemma 4 31B about 15% cheaper for the same work.

How do you compare the two?

We use the 36 benchmark tests both models have published scores on. The verdict counts the 14 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 22 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.