Gemma 3 27B vs Llama 3.1 70B
Change modelsGemma 3 27BGoogle
Llama 3.1 70BMeta
IndexAll categories, out of 100.
25.3 ±4.5 best
24.8 ±5.3
GapIndex points below the top.
Top
−0.5 · no clear gap
CoverageIndex weight that has results.
100% best
90%
SpeedOutput tokens per second.
31/s
—
EloLMArena rating and rank.
—
—
ReleasedMonth of release.
2025-03
2024-07
Input / 1MUS dollars per 1M input tokens.
—
—
Output / 1MUS dollars per 1M output tokens.
—
—
1K chats2K tokens in, 500 out, each.
—
—
1K documents20K tokens in, 1K out, each.
—
—
1K agent tasks200K tokens in, 20K out, each.
—
—
ContextMaximum tokens in one request.
32K
128K best
AgenticMulti-step tasks with tools.
21.9 best
17.1
CodingCode writing and repair.
26.2 best
21.4
ReasoningLogic problems and puzzles.
24.6 best
23.7
MultimodalTasks with images and text.
17.6
—
KnowledgeFacts and expert knowledge.
24.9
27.6 best
MultilingualTasks in many languages.
43.7 best
41.0
InstructionTasks with strict prompt rules.
18.7
27.8 best
MathMath problems.
28.5 best
13.9
Benchmarks
13 sharedOnly Gemma 3 27B: 6Only Llama 3.1 70B: 7
Gemma 3 27B
Llama 3.1 70B