メインコンテンツへスキップ

Claude Haiku 5.5 (Max)

AnthropicClaudeProprietary

説明

Claude Haiku 5.5 is Anthropic's cheapest, fastest, and most capable small model to date, and the successor to Claude Haiku 4.5. It targets high-volume, cost-sensitive work (summaries, compaction, classification, database queries) and pairs with Opus 5.5 / Sonnet 5.5 as a coding subagent; it is especially strong for speed-sensitive tasks like live support and browser use. It is the first Haiku-class model with adjustable effort (low/medium/high/xhigh/max; Claude Platform default medium). It accepts text and image input (including file/PDF) with text output, supports multilingual and vision workloads, tool use, and adaptive thinking. Context window is 1M tokens with 128K max output; knowledge cutoff is June 2026. Claude Platform pricing is tiered by prompt length: prompts up to 100K tokens are $0.10/$0.50 per million input/output (cache read $0.01, 5m cache write $0.125); prompts over 100K are $0.50/$2.50 (cache read $0.05, 5m cache write $0.625). Catalog provider pricing stores the <=100K tier. Capability scores are self-reported from the Claude Haiku 5.5 System Card (Table 8.1.A and §8) and the launch announcement. Available on the Claude API as `claude-haiku-5-5`.

リリース日
2026-10-07
パラメータ
—
コンテキスト長
1.0M
モダリティ
image, pdf, text

能力レーダー

44
general
55
coding
50
reasoning
51
science
30
agents
60
multimodal

ランキング

ベンチマークスコア (LLM Stats)

(LLM Stats (zeroeval))

Agents

AA-Briefcase v1.11578.00 / 3000自己申告
Program Bench82.0%自己申告
OfficeQA73.5%自己申告
OfficeQA Pro60.3%自己申告

Biology

LatchBio SpatialBench Verified67.7%自己申告
Protocols Understanding (V2)64.1%自己申告
Protocols Troubleshooting58.0%自己申告
Medicinal Chemistry (ADME)58.0%自己申告
Protein Design — Library Ranking45.3%自己申告
Biomedical image analysis43.5%自己申告
Protein Design — Sequence Generation33.1%自己申告
Morphology-to-molecule matching25.4%自己申告

Code

BenchCAD (with Python tool)87.0%自己申告
BenchCAD67.0%自己申告
FrontierSWE V243.8%自己申告

General

GDPval-AA 2.11620.00 / 3000自己申告
Global-MMLU87.8%自己申告

Healthcare

HealthBench Professional (raw)71.0%自己申告
HealthBench Professional64.8%自己申告
HealthBench61.6%自己申告
PhysicianBench43.0%自己申告

Language

MILU87.6%自己申告

Multimodal

OSWorld 2.1 (partial)72.4%自己申告
OSWorld 2.1 (strict)37.1%自己申告

Reasoning

SWE-bench Multilingual83.7%自己申告
SWE-Bench ProPrinceton NLP (2024)64.8%自己申告
FrontierCode 1.1 (Extended)58.4%自己申告
Humanity's Last Exam (with tools)57.4%自己申告
FrontierCode 1.146.4%自己申告
Humanity's Last Exam (no tools, text-only)45.9%自己申告
Terminal-Bench 4.039.2%自己申告
SWE-Bench Multimodal30.7%自己申告
Terminal-Bench-Science 0.120.6%自己申告

Science

LatchBio SingleCellBench56.2%自己申告

Vision

Chartography86.2%自己申告
Chartography (no tools)46.4%自己申告

AA評価指数

(Artificial Analysis)
Lcr(Artificial Analysis)
82.7
Scicode(UIUC + Argonne National Lab (2024))
55.0
Hle(Center for AI Safety + Scale AI (2025))
44.4
Intelligence Index(Artificial Analysis)
43.4
Terminalbench V4 0
32.8

LLM Statsカテゴリスコア

(LLM Stats (zeroeval))
Productivity
100
Agents
100
Reasoning
100
General
100
Language
90
Multimodal
60
Science
60
Healthcare
60
Biology
60
Code
60
Vision
60
Math
50
Tool Calling
30

価格設定

入力価格$0.1 / 1Mトークン
出力価格$0.5 / 1Mトークン
混合価格(3:1)$0.2 / 1Mトークン
キャッシュ読み取り価格$0.01 / 1Mトークン
キャッシュ書き込み価格$0.125 / 1Mトークン

速度

トークン/秒227.9
初トークン遅延292.43s
初回答遅延292.43s

プロバイダー価格ランキング

プロバイダー価格ランキング

1 プロバイダー

プロバイダー入力出力
1Anthropic
$0
$0

このモデルの異なるAPIプロバイダー間の価格を比較。

外部リンク