Claude 3.5 Sonnet vs Claude 3 Haiku
Change modelsClaude 3.5 SonnetAnthropic
Claude 3 HaikuAnthropic
IndexAll categories, out of 100.
32.2 ±9.2 best
22.0 ±5.4
GapIndex points below the top.
Top
−10.2 · no clear gap
CoverageIndex weight that has results.
55%
95% best
SpeedOutput tokens per second.
—
73/s
EloLMArena rating and rank.
1298 best
1195
ReleasedMonth of release.
2024-06
2024-03
Input / 1MUS dollars per 1M input tokens.
$3.00
$0.25 best
Output / 1MUS dollars per 1M output tokens.
$15.00
$1.25 best
1K chats2K tokens in, 500 out, each.
$13.50
$1.13 best
1K documents20K tokens in, 1K out, each.
$75.00
$6.25 best
1K agent tasks200K tokens in, 20K out, each.
$900.00
$75.00 best
ContextMaximum tokens in one request.
200K
200K
AgenticMulti-step tasks with tools.
—
16.8
CodingCode writing and repair.
34.0 best
31.2
ReasoningLogic problems and puzzles.
—
31.6
MultimodalTasks with images and text.
44.7 best
5.0
KnowledgeFacts and expert knowledge.
37.4 best
24.1
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
—
22.8
MathMath problems.
25.2 best
8.1
Benchmarks
4 sharedOnly Claude 3.5 Sonnet: 23Only Claude 3 Haiku: 8
Claude 3.5 Sonnet
Claude 3 Haiku