Llama 4 Scout vs Phi-4-reasoning
Change modelsLlama 4 ScoutMeta
Phi-4-reasoningMicrosoft
IndexAll categories, out of 100.
26.8 ±4.8
27.1 ±6.4 best
GapIndex points below the top.
−0.3 · no clear gap
Top
CoverageIndex weight that has results.
95% best
85%
SpeedOutput tokens per second.
63/s best
41/s
EloLMArena rating and rank.
—
—
ReleasedMonth of release.
2025-04
2025-04
Input / 1MUS dollars per 1M input tokens.
$0.10
—
Output / 1MUS dollars per 1M output tokens.
$0.30
—
1K chats2K tokens in, 500 out, each.
$0.35
—
1K documents20K tokens in, 1K out, each.
$2.30
—
1K agent tasks200K tokens in, 20K out, each.
$26.00
—
ContextMaximum tokens in one request.
1.31M best
32K
AgenticMulti-step tasks with tools.
23.6 best
5.0
CodingCode writing and repair.
23.6
34.0 best
ReasoningLogic problems and puzzles.
32.4 best
22.1
MultimodalTasks with images and text.
24.0
—
KnowledgeFacts and expert knowledge.
28.0
33.3 best
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
25.9 best
21.1
MathMath problems.
24.2
35.5 best
Benchmarks
6 sharedOnly Llama 4 Scout: 12Only Phi-4-reasoning: 6
Llama 4 Scout
Phi-4-reasoning