Compare / Nova Lite 1.0 vs Gemma 3 4B

Nova Lite 1.0 vs Gemma 3 4B

Which is cheaper, which is more capable, and which does better at each kind of work, from 15 results on 3 sources that measured both models.

AmazonNova Lite 1.0 Released 3 Dec 2024 · $0.060 in · $0.24 out
GoogleGemma 3 4B Released 11 Mar 2025 · $0.040 in · $0.050 out
Nova Lite 1.0 wins 72 too close to callGemma 3 4B wins 6

Of the 15 results both models have, the better value on each. Results whose reported ranges overlap are too close to call.

Results as of 19 November 2026 (catalogue release 2026-11-19-3608133e8331). See also Amazon vs Google, every model from both labs.

Nova Lite 1.0 or Gemma 3 4B? The short answer

They score about the same, so pick Gemma 3 4B: it costs 2.5 times less per token.

  • Gemma 3 4B is cheaper: $0.042 against $0.10 per million tokens, 2.5 times less.
  • Nova Lite 1.0 scores higher on the Artificial Analysis Intelligence Index: 6.7 against 4.8.
  • Gemma 3 4B is newer: released 11 Mar 2025, against 3 Dec 2024 for Nova Lite 1.0.
  • On third-party benchmarks, on the results both models have: Nova Lite 1.0 does better for reasoning and knowledge and sticking to the facts; Gemma 3 4B for writing.

Head to head

6.7
Intelligence Index
4.8
$0.10
Price per million tokens, blendedlower is better
$0.042
300,000
Context window, tokens
131,072

Every fact side by side

Nova Lite 1.0Gemma 3 4B
Price per million tokens$0.060 in · $0.24 out$0.040 in · $0.050 out
Blended (3 in : 1 out)$0.10$0.042
Cheapest host, blended$0.10 (Amazon Bedrock)$0.042 (Parasail, BF16)
Intelligence Index6.74.8
Spring Prompt overall––
Output speed, tokens a second––
Context window300,000 tokens131,072 tokens
Released3 Dec 202411 Mar 2025 (newer)
WeightsClosedOpen
Developer based inthe USthe US

Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.

Which is better for what

Each model's rank among the models on sale today, for each kind of work, from our own benchmarks and licensed sources. Further right is better; the better rank is in green.

Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls not measured 158 of 159
Reasoning and knowledgeHard questions across science, maths and general knowledge 170 of 187 181 of 187
WritingWhat people prefer in blind comparisons, and judged writing quality 163 of 169 157 of 169
Sticking to the factsSummarising without inventing things, and factual answers 24 of 139 32 of 139

Every shared result

Each model's best result on every measure both have, grouped by benchmark, with the setting that produced it. Under each measure, every model that benchmark has measured is a grey tick, best to the right, so you can see where the two sit in the field. The headline measures of each benchmark come first; the rest are under "Show all". Green marks the better value; where a source reports ranges that overlap, neither is marked.

MeasureNova Lite 1.0Gemma 3 4B
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Overall · TextArena rating, higher is better 289 · 255 of 387 measured 1,260 1,303
Business, management and finance · TextArena rating, higher is better 273 · 208 of 380 measured 1,269 1,330
Creative writing · TextArena rating, higher is better 285 · 243 of 385 measured 1,221 1,275
Artificial Analysis reported by Artificial Analysis · 3 results
Artificial Analysis Intelligence Indexindex score, higher is better 225 · 246 of 249 measured 6.7 4.8
GPQA Diamond% of questions, higher is better 213 · 227 of 232 measured 43.3% 29.1%
Humanity's Last Exam% of questions, higher is better 217 · 194 of 250 measured 4.3% 5.3%
Vectara Hallucination Leaderboard reported by Vectara Hallucination Leaderboard · 2 results
Hallucination rate · Vectara% of summaries, lower is better 25 · 28 of 104 measured 6.1% 6.4%
Answer rate · Vectara% of documents, higher is better 17 · 103 of 104 measured 99.9% 67.3%
Show all 15 results (7 more)
MeasureNova Lite 1.0Gemma 3 4B
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Expert prompts · TextArena rating, higher is better 266 · 270 of 338 measured 1,262 1,260
Instruction following · TextArena rating, higher is better 288 · 270 of 387 measured 1,244 1,267
Writing, literature and language · TextArena rating, higher is better 286 · 237 of 386 measured 1,240 1,298
Artificial Analysis reported by Artificial Analysis · 4 results
IFBench% of instructions, higher is better 163 · 187 of 195 measured 34.2% 28.3%
Long-context reasoning (AA-LCR)% of questions, higher is better 213 · 226 of 241 measured 19.0% 6.7%
Terminal-Bench Hard% of tasks, higher is better 180 · 180 of 190 measured 0.8% 0.8%
τ²-bench telecom% of tasks, higher is better 172 · 187 of 193 measured 17.5% 5.0%

Quick answers

Is Nova Lite 1.0 better than Gemma 3 4B?

They score about the same, so pick Gemma 3 4B: it costs 2.5 times less per token.

Which is cheaper, Nova Lite 1.0 or Gemma 3 4B?

Gemma 3 4B is cheaper: $0.042 against $0.10 per million tokens, 2.5 times less.

Which is better for reasoning and knowledge, Nova Lite 1.0 or Gemma 3 4B?

Nova Lite 1.0 ranks 170 of 187 models on sale for hard questions across science, maths and general knowledge, against 181 for Gemma 3 4B.

Which is better for writing, Nova Lite 1.0 or Gemma 3 4B?

Gemma 3 4B ranks 157 of 169 models on sale for what people prefer in blind comparisons, and judged writing quality, against 163 for Nova Lite 1.0.

Which is better for sticking to the facts, Nova Lite 1.0 or Gemma 3 4B?

Nova Lite 1.0 ranks 24 of 139 models on sale for summarising without inventing things, and factual answers, against 32 for Gemma 3 4B.

Which has the bigger context window, Nova Lite 1.0 or Gemma 3 4B?

Nova Lite 1.0: 300,000 tokens, against 131,072 for Gemma 3 4B.

Run Nova Lite 1.0 and Gemma 3 4B on your own prompt.

Benchmarks aren't your task. Test both side by side on your prompt, with every answer checked against your rules and the cost of each one.