Перейти к основному содержанию

Claude Opus 4.7 (Max)

AnthropicClaudeProprietary

Описание

Claude Opus 4.7 is Anthropic's latest Opus-class model, a direct upgrade to Opus 4.6 with notable improvements in advanced software engineering, particularly on the most difficult tasks. It handles complex, long-running agentic workflows with rigor and consistency, follows instructions more literally and precisely, and verifies its own outputs before reporting back. Substantially improved vision supports high-resolution images up to 2,576 pixels on the long edge (~3.75 megapixels, over 3x prior Claude models), unlocking dense screenshot reading, complex diagram extraction, and pixel-perfect references. Better file system-based memory enables coherent multi-session work. Introduces a new 'xhigh' effort level between 'high' and 'max' for finer control over the reasoning/latency tradeoff, and ships with task budgets (public beta) on the Claude Platform. Uses an updated tokenizer (inputs may map to ~1.0-1.35x more tokens than Opus 4.6). Released with automated safeguards that detect and block prohibited or high-risk cybersecurity uses. Available across Claude products, the Claude API, Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry. Pricing: $5/$25 per million tokens (input/output), unchanged from Opus 4.6.

Дата выхода
2026-04-16
Параметры
—
Длина контекста
1.0M
Модальности
image, pdf, text

Радар способностей

60
general
70
coding
70
reasoning
77
scienceоцен.
70
agents
80
multimodal

При отсутствии специализированных научных бенчмарков Science оценивается по научным категориям LLM Stats или на основе рассуждений.

Рейтинги

Оценки бенчмарков (LLM Stats)

(LLM Stats (zeroeval))

Agents

Finance Agent64.4%Сам.
Finance Agent v251.5%

Code

CyberGym73.1%Сам.
FrontierSWE63.0%

Language

MMMLU91.5%Сам.

Legal

Legal Agent Benchmark7.1%

Math

LiveBench76.9%

Multimodal

OSWorld-Verified78.0%Сам.

Reasoning

GPQANYU + Cohere + Anthropic (2023)94.2%Сам.
CharXiv-R91.0%Сам.
SWE-Bench Verified87.6%Сам.
BrowseCompOpenAI (2025)79.3%Сам.
MCP Atlas77.3%Сам.
Terminal-Bench 2.0Stanford × Laude Institute (2026)69.4%Сам.
SWE-Bench ProPrinceton NLP (2024)64.3%Сам.
Humanity's Last Exam54.7%Сам.
FrontierCode 1.138.5%

Индексы оценки AA

(Artificial Analysis)

Нет данных AA оценки

Оценки категорий LLM Stats

(LLM Stats (zeroeval))
Language
90
Physics
90
Frontend Development
90
Biology
90
Chemistry
90
Multimodal
80
Search
80
Long Context
70
Math
70
Reasoning
70
Safety
70
Code
70
Tool Calling
70
Vision
70
Finance
60
General
60
Agents
60
Healthcare
30
Legal
10

Цены

Цена ввода$5 / 1M токенов
Цена вывода$25 / 1M токенов
Смешанная цена (3:1)$10 / 1M токенов
Цена чтения кэша$0.5 / 1M токенов
Цена записи кэша$6.25 / 1M токенов

Скорость

Токенов/сек47.8
Задержка первого токена23.02s
Время до первого ответа23.02s

Рейтинг цен провайдеров

Рейтинг цен провайдеров

23 провайдеров

Самый дешевый: JZSTOKENСамый дорогой: Kenari
ПровайдерВводВывод
1JZSTOKENСамый дешевый
$1.17
$5.84
2xxapi
$1.923
$9.615
3KoRouter
$3.25
$16.25
4Kunavo
$3.5
$17.5
5AnthropicОсновной
$5
$25
6DeepInfra
$5
$25
7302.AI
$5
$25
8Abacus
$5
$25
9OpenCode
$5
$25
10FrogBot
$5
$25
11Venice AI
$5
$25
12AIHubMix
$5
$25
13Azure Cognitive Services
$5
$25
14Requesty
$5
$25
15DevPass (LLM Gateway)
$5
$25
16Azure
$5
$25
17Auriko
$5
$25
18Snowflake Cortex
$5
$25
19FreeModel
$5
$25
20Neon
$5
$25
21CommandCode
$5
$25
22Opper
$5.5
$27.5
23Kenari
$5.71
$29.71

Сравнение цен разных API-провайдеров для этой модели.

Внешние ссылки