Llama 3.1 405B vs Phi-4-reasoning
Change modelsLlama 3.1 405BMeta
Phi-4-reasoningMicrosoft
IndexAll categories, out of 100.
27.0 ±8.9
27.1 ±6.4 best
GapIndex points below the top.
−0.1 · no clear gap
Top
CoverageIndex weight that has results.
65%
85% best
SpeedOutput tokens per second.
29/s
41/s best
EloLMArena rating and rank.
—
—
ReleasedMonth of release.
2024-07
2025-04
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.
128K best
32K
AgenticMulti-step tasks with tools.
15.4 best
5.0
CodingCode writing and repair.
—
34.0
ReasoningLogic problems and puzzles.
30.8 best
22.1
MultimodalTasks with images and text.
—
—
KnowledgeFacts and expert knowledge.
26.3
33.3 best
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
33.5 best
21.1
MathMath problems.
18.9
35.5 best
Benchmarks
7 sharedOnly Llama 3.1 405B: 3Only Phi-4-reasoning: 5
Llama 3.1 405B
Phi-4-reasoning