Compare / Gemma 3 4B vs Llama 3.1 8B Instruct

Gemma 3 4B vs Llama 3.1 8B Instruct

Which is cheaper, which is more capable, and which does better at each kind of work, from 26 results on 4 sources that measured both models.

GoogleGemma 3 4B Released 11 Mar 2025 · $0.040 in · $0.050 out
MetaLlama 3.1 8B Instruct Released 23 Jul 2024 · $0.050 in · $0.080 out
Gemma 3 4B wins 102 too close to callLlama 3.1 8B Instruct wins 14

Of the 26 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 3.1 8B Instruct? The short answer

Pick Llama 3.1 8B Instruct if quality matters most: it scores higher (Intelligence Index 6.9 against 4.8). Pick Gemma 3 4B if cost matters more: it costs 26% less per token.

  • Gemma 3 4B is cheaper: $0.042 against $0.058 per million tokens, 26% less.
  • Llama 3.1 8B Instruct scores higher on the Artificial Analysis Intelligence Index: 6.9 against 4.8.
  • Gemma 3 4B is newer: released 11 Mar 2025, against 23 Jul 2024 for Llama 3.1 8B Instruct.
  • On third-party benchmarks, on the results both models have: Llama 3.1 8B Instruct does better for agents and tool use.

Head to head

4.8
Intelligence Index
6.9
$0.042
Price per million tokens, blendedlower is better
$0.058
131,072
Context window, tokens
131,072

Every fact side by side

Gemma 3 4BLlama 3.1 8B Instruct
Price per million tokens$0.040 in · $0.050 out$0.050 in · $0.080 out
Blended (3 in : 1 out)$0.042$0.058
Cheapest host, blended$0.042 (Parasail, BF16)$0.025 (DeepInfra, FP8)
Intelligence Index4.86.9
Spring Prompt overall––
Output speed, tokens a second––
Context window131,072 tokens131,072 tokens
Released11 Mar 2025 (newer)23 Jul 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 159 153 of 159
Reasoning and knowledgeHard questions across science, maths and general knowledge 181 of 187 166 of 187
WritingWhat people prefer in blind comparisons, and judged writing quality 157 of 169 164 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 3.1 8B Instruct
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Overall · TextArena rating, higher is better 255 · 314 of 387 measured 1,303 1,211
Business, management and finance · TextArena rating, higher is better 208 · 308 of 380 measured 1,330 1,210
Creative writing · TextArena rating, higher is better 243 · 313 of 385 measured 1,275 1,177
Artificial Analysis reported by Artificial Analysis · 3 results
Artificial Analysis Intelligence Indexindex score, higher is better 246 · 222 of 249 measured 4.8 6.9
GPQA Diamond% of questions, higher is better 227 · 228 of 232 measured 29.1% 25.9%
Humanity's Last Exam% of questions, higher is better 194 · 194 of 250 measured 5.3% 5.3%
Berkeley Function Calling Leaderboard (BFCL) V4 reported by Berkeley Function Calling Leaderboard (BFCL) V4 · 3 results
Overall accuracy · BFCL% correct, higher is better 74 · 63 of 81 measured 19.6%bfcl prompt 25.8%bfcl prompt
Irrelevance detection · BFCL% correct, higher is better 71 · 76 of 81 measured 53.9%bfcl prompt 42.7%bfcl prompt
Memory · BFCL% correct, higher is better 53 · 51 of 81 measured 8.6%bfcl prompt 10.8%bfcl prompt
UGI Leaderboard reported by UGI Leaderboard · 3 results
Writing score · UGIscore out of 100, higher is better 185 · 181 of 230 measured 24.0 24.8
Requested-length error · UGI% off the requested word count, lower is better 33 · 111 of 230 measured 9.0% 20.0%
Style adherence · UGIscore from 0 to 1, higher is better 214 · 122 of 230 measured 0.29 0.34
Show all 26 results (14 more)
MeasureGemma 3 4BLlama 3.1 8B Instruct
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Expert prompts · TextArena rating, higher is better 270 · 304 of 338 measured 1,260 1,198
Instruction following · TextArena rating, higher is better 270 · 315 of 387 measured 1,267 1,191
Writing, literature and language · TextArena rating, higher is better 237 · 315 of 386 measured 1,298 1,182
Artificial Analysis reported by Artificial Analysis · 6 results
Terminal-Bench 2.1% of tasks, higher is better 127 · 125 of 131 measured 0.4% 1.5%
Artificial Analysis Coding Indexindex score, higher is better 129 · 126 of 130 measured 2.7 5.4
IFBench% of instructions, higher is better 187 · 186 of 195 measured 28.3% 28.6%
Long-context reasoning (AA-LCR)% of questions, higher is better 226 · 215 of 241 measured 6.7% 18.0%
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 187 · 174 of 193 measured 5.0% 16.4%
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 · 48 of 81 measured 0.4%bfcl prompt 11.1%bfcl prompt
Relevance detection · BFCL% correct, higher is better 1 · 8 of 81 measured 100.0%bfcl prompt 93.8%bfcl prompt
Single-turn calls (curated) · BFCL% correct, higher is better 74 · 43 of 81 measured 61.1%bfcl prompt 84.0%bfcl prompt
Single-turn calls (user-contributed) · BFCL% correct, higher is better 64 · 43 of 81 measured 60.8%bfcl prompt 70.8%bfcl prompt
Web search · BFCL% correct, higher is better 56 · 47 of 81 measured 1.0%bfcl prompt 3.0%bfcl prompt

Quick answers

Is Gemma 3 4B better than Llama 3.1 8B Instruct?

Pick Llama 3.1 8B Instruct if quality matters most: it scores higher (Intelligence Index 6.9 against 4.8). Pick Gemma 3 4B if cost matters more: it costs 26% less per token.

Which is cheaper, Gemma 3 4B or Llama 3.1 8B Instruct?

Gemma 3 4B is cheaper: $0.042 against $0.058 per million tokens, 26% less.

Which is better for agents and tool use, Gemma 3 4B or Llama 3.1 8B Instruct?

Llama 3.1 8B Instruct ranks 153 of 159 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 158 for Gemma 3 4B.

Which is better for reasoning and knowledge, Gemma 3 4B or Llama 3.1 8B Instruct?

Llama 3.1 8B Instruct ranks 166 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, Gemma 3 4B or Llama 3.1 8B Instruct?

Gemma 3 4B ranks 157 of 169 models on sale for what people prefer in blind comparisons, and judged writing quality, against 164 for Llama 3.1 8B Instruct.

Run Gemma 3 4B and Llama 3.1 8B 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.