LANGUAGE MODEL May 2025 Anthropic Last updated: Apr 29, 2026

Claude 4

Opus & Sonnet (Agentic Frontier Family)

Anthropic's first model family explicitly designed for sustained autonomous work — running for hours on multi-step tasks without human intervention. Opus 4 became the dominant model behind "AI software engineer" products in 2025, while Sonnet 4 stayed the workhorse mid-tier. Both shipped with native tool use, file editing, and computer-use improvements.

Try Claude API Docs ↗

Intelligence

Medium

Speed

Slow

39 tok/s output

Cost

High

$18.75 in / $75.00 out

Context

200K

Up to 200,000 tokens

How are Intelligence, Speed & Cost bucketed?

Intelligence and Speed buckets are percentile ranks on Artificial Analysis. Cost buckets are fixed dollar thresholds keyed off output-token price ($/M out).

Intelligence

Top 1%≤ 1%
Top 5%≤ 5%
Top 10%≤ 10%
Good≤ 25%
Medium≤ 50%
Below avg> 50%

Speed

Top 1%≥ 345 tok/s
Top 5%≥ 237 tok/s
Top 10%≥ 196 tok/s
Good≥ 146 tok/s
Medium≥ 90 tok/s
Slow< 90 tok/s

Cost

Freeopen weights · self-host
Low< $1 / M out
Moderate$1–5 / M out
High≥ $5 / M out

Official ↗ Artificial Analysis ↗

Why it matters

Claude 4 is the model that made "have AI write code, autonomously, and you review the PR" a normal workflow at frontier engineering teams. The labor-market implications — entry-level engineering hiring contraction visible in Q3 2025 BLS data — are downstream of this and similar releases.

Core Capabilities

Long Documents

Handles entire codebases, books, and multi-doc RAG.

Multimodal

Combines text, vision, and audio in one model.

Generative

Produces images, video, audio, or other media.

Agent Workflows

Built for tool use and autonomous tasks.

Context Window

200k tokens

≈ 154+ pages

4k Chat 聊天

32k Long docs 长文档

128k Books 整本书

400k Multi-doc 多文档

1M Codebase 整个代码库

10M

200k

Availability

API

Available

Product / App

Available

Open Source

Not released

Enterprise

Available

Pricing Model

Pay per token

Input and output billed separately.

Pay-per-token

Capability / Performance

Where this model sits relative to the middle 60% of models in the tree. All scores are 0–10 (higher is better).

Lower 20% Upper 80% This model

Reasoning

AA Intelligence Index · scaled to 10

1.7

5.6

Coding

SciCode · scaled to 10

1.8

4.3

4.7

Agentic tasks

Terminal-Bench Hard · scaled to 10

0.2

3.6

5.3

Context / memory

Context window size · log-scaled

6.0

9.0

6.7

Cost efficiency

Input price ($/M tokens) · cheaper scores higher

6.2

10.0

0.2

Lower 20% 20th percentile — 20% of models score below this This model Where the current model lands Upper 80% 80th percentile — only 20% of models score above this Percentile boundaries are computed across every model in the tree that reports the underlying benchmark for each capability.

What it feels like

Sustains hours-long agent loops without losing the thread — testers ran Opus 4 on 7+ hour autonomous coding sessions
Best-in-class for real-world software engineering at release; SWE-bench Verified gains over Claude 3.7 Sonnet
Native MCP support makes tool-using agents noticeably more reliable in production
File editing and computer use feel meaningfully better than 3.7 — fewer 'almost works' debugging cycles
Opus 4 is slower and more expensive than alternatives; reserve for the hardest tasks
Sonnet 4 is the workhorse — same agentic strengths at a much friendlier price

Reviews: Anthropic — Claude 4 announcement ↗ · 16x Engineer — Claude 4 evaluation results ↗ · DataCamp — Claude 4 review ↗

Best use cases

AI-software-engineer products (Devin, Cursor, Cline) that need long-horizon autonomy
Agentic workflows with tool use, file editing, and computer use
Complex bug fixes that span multiple files / systems
Multi-hour research or coding sessions where context retention matters