Inference
EN
Anthropic

Claude 3 Haiku

Claude 3 Haiku is a standard model from Anthropic. 11 benchmarks count toward its index, in 7 categories.

  • Closed weights
  • Standard
  • Input: text · image
  • Released 2024-03
RankedEnough results for an index.
IndexAll categories, out of 100.22.0±5.4Median 51.0
CoverageIndex weight that has results.95%
SpeedOutput tokens per second.73/sMedian 67/s
First tokenSeconds to the first output token.0.4 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.$0.25Median $0.60
Output / 1MUS dollars per 1M output tokens.$1.25Median $2.55
ContextMaximum tokens in one request.200KMedian 262K
EloLMArena rating and rank.1195±4 #326Median 1422

Details

From OpenRouter
API IDThe model ID on OpenRouter.
anthropic/claude-3-haiku
Max outputMaximum output tokens per request.
4K
Knowledge cutoffLast date of training data.
31 Aug 2023
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
No
Cache read / 1MUS dollars per 1M cached tokens.
$0.03
WeightsLink to the published weights.
—
RetiresDate OpenRouter removes the model.
—

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.1 counted16.8—
CodingCode writing and repair.1 counted31.2—
ReasoningLogic problems and puzzles.2 counted31.6202 / 227
MultimodalTasks with images and text.1 counted5.0—
KnowledgeFacts and expert knowledge.3 counted24.1261 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.1 counted22.8—
MathMath problems.2 counted8.1168 / 168

Results

11 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
Critical Physics TasksReasoning0.0%150 / 214
SWE-bench Verified · SWE-bench/experimentsCoding40.6%49 / 65
Artificial Analysis IFBenchInstruction36.1%121 / 146
GPQA diamond · Epoch AIKnowledge36.3% ±2.4182 / 217
Artificial Analysis Long Context ReasoningReasoning27.7%184 / 217
MATH level 5Math14.9% ±0.775 / 88
τ²-Bench Tool-Agent-User EvaluationAgentic21.1%129 / 149
Artificial Analysis Omniscience AccuracyKnowledge17.6%142 / 164
Artificial Analysis Humanity's Last ExamKnowledge4.1%199 / 222
OTIS Mock AIME 2024-2025Math1.8% ±0.8187 / 200
Artificial Analysis GPQA DiamondKnowledge37.4%192 / 199
Artificial Analysis MMMU-ProMultimodal30.8%111 / 113

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reuseSWE-bench teamNo licence statedEpoch AICC BY

History

5 releases
Index22.6 → 22.0
23 Oct 202612 Nov 2026
Output / 1M$1.25 → $1.25
23 Oct 202612 Nov 2026

Questions

How much does Claude 3 Haiku cost?
$0.25 per 1M input tokens, $1.25 per 1M output tokens.
How good is Claude 3 Haiku?
Its index is 22.0 out of 100. That is place 276 of 283 ranked models.
What is Claude 3 Haiku best at?
Reasoning: 31.6 out of 100, place 202 of 227.
How fast is Claude 3 Haiku?
73 output tokens per second, 0.4 s to the first token.
What is the context window of Claude 3 Haiku?
200K tokens.
Is Claude 3 Haiku open source?
No. Its weights are not published.
When was Claude 3 Haiku released?
March 2024.

Badge

Claude 3 Haiku on Inference360: place 276 of 283, index 22.0
HTML
<a href="https://inference360.ai/models/claude-3-haiku"><img src="https://inference360.ai/badges/claude-3-haiku.svg" alt="Claude 3 Haiku on Inference360: place 276 of 283, index 22.0" height="20"></a>
Markdown
[![Claude 3 Haiku on Inference360: place 276 of 283, index 22.0](https://inference360.ai/badges/claude-3-haiku.svg)](https://inference360.ai/models/claude-3-haiku)

Compare with

Gemma 4 E2B21.9Phi-422.4Claude 3 Opus23.3Gemini 1.5 Flash20.2Claude 3.5 Sonnet32.2

Lower price

dots3-note Preview63.4 · FreeApodex 1.1 Mini55.7 · FreeNemotron 3 Ultra50.8 · Free

More from Anthropic

All models
Claude 3 Opus23.3Claude 3.5 Sonnet32.2Claude 3.5 Haiku24.1Claude 3.7 Sonnet Thinking41.4Claude 3.7 Sonnet37.7