Claude Opus 4 vs Claude Haiku 4.5 Thinking
Change modelsClaude Opus 4Anthropic
Claude Haiku 4.5 ThinkingAnthropic
IndexAll categories, out of 100.
43.9 ±5.3 best
43.5 ±7.0
GapIndex points below the top.
Top
−0.4 · no clear gap
CoverageIndex weight that has results.
90% best
75%
SpeedOutput tokens per second.
—
—
EloLMArena rating and rank.
1366
—
ReleasedMonth of release.
—
2025-10
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.
—
200K
AgenticMulti-step tasks with tools.
49.6
50.2 best
CodingCode writing and repair.
47.3 best
42.5
ReasoningLogic problems and puzzles.
38.6
—
MultimodalTasks with images and text.
50.6 best
30.1
KnowledgeFacts and expert knowledge.
48.2 best
43.6
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
—
—
MathMath problems.
39.4
51.6 best
Benchmarks
10 sharedOnly Claude Opus 4: 15Only Claude Haiku 4.5 Thinking: 18
Claude Opus 4
Claude Haiku 4.5 Thinking