Breaking

Anthropic Released Claude Opus 4.7 With Adaptive Thinking for Coding Agents

October 3, 2026 at 22:00 EDT

Foundation Models
AI Agents
Software Dev & Coding

Anthropic released Claude Opus 4.7 on April 16, 2026, positioning the model as a generally available upgrade for advanced coding, enterprise workflows and long-running agentic tasks. The launch centers on adaptive thinking, higher-resolution vision and stronger instruction following while keeping Opus 4.6 pricing unchanged.

Apr 16, 2026 | Anthropic

Claude Opus 4.7 targets long-running coding agents

The release pairs adaptive thinking with a 1M-token context window, stronger vision and migration work for teams moving from Opus 4.6.

context tokens

128K

max output tokens

$5/$25

input/output per 1M tokens

2,576

pixels on image long edge

Visual acuity benchmark

A reported computer-use visual test showed Opus 4.7 nearly doubling Opus 4.6.

54.5%

Opus 4.6

98.5%

Opus 4.7

How adaptive thinking changes the run

PromptTask and constraints

ReasoningEffort chosen per step

WorkTools, code and review

Fixed thinking budgets give way to effort controls and task budgets for longer agent runs.

Where teams see lift

Long-running coding tasks
Code review and debugging
Dense screenshots and visual workflows

Where teams must tune

1.0-1.35x tokenizer shift
No fixed thinking budget
Prompt and harness retuning

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.