Compare / Gemma 3 4B vs Llama 3.2 1B Instruct

Gemma 3 4B vs Llama 3.2 1B Instruct

Which is cheaper, which is more capable, and which does better at each kind of work, from 24 results on 4 sources that measured both models.

GoogleGemma 3 4B Released 11 Mar 2025 · $0.040 in · $0.050 out
MetaLlama 3.2 1B Instruct Released 25 Sep 2024 · $0.027 in · $0.20 out
Gemma 3 4B wins 202 too close to callLlama 3.2 1B Instruct wins 2

Of the 24 results both models have, the better value on each. Results whose reported ranges overlap are too close to call.

Results as of 16 November 2026 (catalogue release 2026-11-16-49bdfa045dee). See also Google vs Meta, every model from both labs.

Gemma 3 4B or Llama 3.2 1B Instruct? The short answer

They score about the same, so pick Gemma 3 4B: it costs 40% less per token.

  • Gemma 3 4B is cheaper: $0.042 against $0.071 per million tokens, 40% less.
  • They score about the same on the Artificial Analysis Intelligence Index (4.8 and 4.8).
  • Gemma 3 4B is newer: released 11 Mar 2025, against 25 Sep 2024 for Llama 3.2 1B Instruct.
  • On third-party benchmarks, on the results both models have: Gemma 3 4B does better for agents and tool use and writing.

Head to head

4.8
Intelligence Index
4.8
$0.042
Price per million tokens, blendedlower is better
$0.071
131,072
Context window, tokens
60,000

Every fact side by side

Gemma 3 4BLlama 3.2 1B Instruct
Price per million tokens$0.040 in · $0.050 out$0.027 in · $0.20 out
Blended (3 in : 1 out)$0.042$0.071
Cheapest host, blended$0.042 (Parasail, BF16)–
Intelligence Index4.84.8
Spring Prompt overall––
Output speed, tokens a second––
Context window131,072 tokens60,000 tokens
Released11 Mar 2025 (newer)25 Sep 2024
WeightsOpenOpen
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 158 of 159 159 of 159
Reasoning and knowledgeHard questions across science, maths and general knowledge 181 of 187 182 of 187
WritingWhat people prefer in blind comparisons, and judged writing quality 157 of 169 168 of 169
Sticking to the factsSummarising without inventing things, and factual answers 32 of 139 not measured

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.

MeasureGemma 3 4BLlama 3.2 1B Instruct
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Overall · TextArena rating, higher is better 255 · 369 of 387 measured 1,303 1,111
Business, management and finance · TextArena rating, higher is better 208 · 367 of 380 measured 1,330 1,070
Creative writing · TextArena rating, higher is better 243 · 366 of 385 measured 1,275 1,083
Artificial Analysis reported by Artificial Analysis · 3 results
Artificial Analysis Intelligence Indexindex score, higher is better 246 · 246 of 249 measured 4.8 4.8
GPQA Diamond% of questions, higher is better 227 · 232 of 232 measured 29.1% 19.6%
Humanity's Last Exam% of questions, higher is better 194 · 192 of 250 measured 5.3% 5.5%
Berkeley Function Calling Leaderboard (BFCL) V4 reported by Berkeley Function Calling Leaderboard (BFCL) V4 · 3 results
Overall accuracy · BFCL% correct, higher is better 74 · 79 of 81 measured 19.6%bfcl prompt 10.8%bfcl fc
Irrelevance detection · BFCL% correct, higher is better 71 · 74 of 81 measured 53.9%bfcl prompt 51.6%bfcl fc
Memory · BFCL% correct, higher is better 53 · 68 of 81 measured 8.6%bfcl prompt 3.2%bfcl fc
UGI Leaderboard reported by UGI Leaderboard · 3 results
Writing score · UGIscore out of 100, higher is better 185 · 223 of 230 measured 24.0 11.8
Requested-length error · UGI% off the requested word count, lower is better 33 · 129 of 230 measured 9.0% 22.0%
Style adherence · UGIscore from 0 to 1, higher is better 214 · 140 of 230 measured 0.29 0.33
Show all 24 results (12 more)
MeasureGemma 3 4BLlama 3.2 1B Instruct
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Expert prompts · TextArena rating, higher is better 270 · 338 of 338 measured 1,260 1,082
Instruction following · TextArena rating, higher is better 270 · 368 of 387 measured 1,267 1,086
Writing, literature and language · TextArena rating, higher is better 237 · 370 of 386 measured 1,298 1,076
Artificial Analysis reported by Artificial Analysis · 4 results
IFBench% of instructions, higher is better 187 · 193 of 195 measured 28.3% 22.8%
Long-context reasoning (AA-LCR)% of questions, higher is better 226 · 226 of 241 measured 6.7% 6.7%
Terminal-Bench Hard% of tasks, higher is better 180 · 185 of 190 measured 0.8% 0.0%
τ²-bench telecom% of tasks, higher is better 187 · 188 of 193 measured 5.0% 0.0%
Berkeley Function Calling Leaderboard (BFCL) V4 reported by Berkeley Function Calling Leaderboard (BFCL) V4 · 5 results
Multi-turn tasks · BFCL% correct, higher is better 75 · 77 of 81 measured 0.4%bfcl prompt 0.0%bfcl fc
Relevance detection · BFCL% correct, higher is better 1 · 77 of 81 measured 100.0%bfcl prompt 43.8%bfcl fc
Single-turn calls (curated) · BFCL% correct, higher is better 74 · 76 of 81 measured 61.1%bfcl prompt 38.4%bfcl fc
Single-turn calls (user-contributed) · BFCL% correct, higher is better 64 · 78 of 81 measured 60.8%bfcl prompt 11.8%bfcl fc
Web search · BFCL% correct, higher is better 56 · 66 of 81 measured 1.0%bfcl prompt 0.0%bfcl fc

Quick answers

Is Gemma 3 4B better than Llama 3.2 1B Instruct?

They score about the same, so pick Gemma 3 4B: it costs 40% less per token.

Which is cheaper, Gemma 3 4B or Llama 3.2 1B Instruct?

Gemma 3 4B is cheaper: $0.042 against $0.071 per million tokens, 40% less.

Which is better for agents and tool use, Gemma 3 4B or Llama 3.2 1B Instruct?

Gemma 3 4B ranks 158 of 159 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 159 for Llama 3.2 1B Instruct.

Which is better for reasoning and knowledge, Gemma 3 4B or Llama 3.2 1B Instruct?

Gemma 3 4B ranks 181 of 187 models on sale for hard questions across science, maths and general knowledge, against 182 for Llama 3.2 1B Instruct.

Which is better for writing, Gemma 3 4B or Llama 3.2 1B Instruct?

Gemma 3 4B ranks 157 of 169 models on sale for what people prefer in blind comparisons, and judged writing quality, against 168 for Llama 3.2 1B Instruct.

Which has the bigger context window, Gemma 3 4B or Llama 3.2 1B Instruct?

Gemma 3 4B: 131,072 tokens, against 60,000 for Llama 3.2 1B Instruct.

Run Gemma 3 4B and Llama 3.2 1B Instruct 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.