Gemini 3.5 Flash vs Gemma 4 31B

Google against Google, the same lab's consecutive models. Prices, scores and speed below are live.

  • Gemma 4 31B costs 93% less on blended price ($0.38 against $5.06 per 1M tokens, 3 input to 1 output).
  • Gemini 3.5 Flash scores higher on the Artificial Analysis Intelligence Index: 33 against 4.4.
  • On the 21 benchmarks both have scores on, Gemini 3.5 Flash leads on 20 and Gemma 4 31B on 1.
  • Gemini 3.5 Flash reads more context: 1M tokens against 262K.
  • Gemini 3.5 Flash is newer: listed May 19, 2026, against Apr 2, 2026.

Side by side

Gemini 3.5 FlashGemma 4 31B
LabGoogleGoogle
ListedMay 19, 2026Apr 2, 2026
Input per 1M tokens$2.25$0.30
Output per 1M tokens$13.50$0.60
Cached input per 1M$0.23-
Blended (3:1) per 1M$5.06$0.38
Context window1M262K
Max output66K16K
Knowledge cutoff2025-01-01-
Intelligence index334.4
Coding index70.143.4
ReasoningYesYes
Reads imagesYesYes
Tool useYesYes
Structured outputYesYes
Measured output speed127 tokens/s-

Every benchmark both are scored on

Each model's best score per benchmark, with its source. Scores from different sources or settings are not always comparable; each links to where it was published.

BenchmarkGemini 3.5 FlashGemma 4 31B
AIME88.9%Epoch AI73.3%Epoch AI
Artificial Analysis Agentic Index26Artificial Analysis and OpenRouter4.2Artificial Analysis and OpenRouter
Artificial Analysis Coding Index70.1Artificial Analysis and OpenRouter43.4Artificial Analysis and OpenRouter
Artificial Analysis Intelligence Index33Artificial Analysis and OpenRouter4.4Artificial Analysis and OpenRouter
Chess Puzzles45%Epoch AI5%Epoch AI
DTBench94.7%Epoch AI82.7%Epoch AI
GPQA Diamond92.3%Artificial Analysis and OpenRouter82%Artificial Analysis and OpenRouter
LMArena (Coding)1493 EloLMArena1459 EloLMArena
LMArena (Overall)1483 EloLMArena1443 EloLMArena
LMArena: Creative Writing1471 EloLMArena1415 EloLMArena
LMArena: Hard Prompts1489 EloLMArena1447 EloLMArena
LMArena: Longer Query1484 EloLMArena1445 EloLMArena
LMArena: Math1494 EloLMArena1464 EloLMArena
LMArena: Multi-Turn1481 EloLMArena1451 EloLMArena
LMCA47.2Epoch AI39.3Epoch AI
SciCode53.9%Epoch AI45.5%Epoch AI
SimpleQA Verified66.2%Epoch AI10.4%Epoch AI
Surface Evolver Bench58.1Epoch AI30.6Epoch AI
tau-bench (Airline)72.7%Artificial Analysis and OpenRouter75.3%Artificial Analysis and OpenRouter
WebDev Arena1506 EloEpoch AI1364 EloEpoch AI
WeirdML62.6%Epoch AI52.3%Epoch AI

More comparisons

Where these figures come from

Prices are GenMagic's own, what the studio and the API charge, recomputed from each provider's price list every few minutes. The indices are the Artificial Analysis indices, and every benchmark score links to its source, re-read daily. Measured speed is GenMagic's own generation records, in aggregate. See Google models, Google models, every LLM's price and our sources.