# Claude Opus 4.8

Latest and most capable model for complex reasoning and agentic coding. Flagship model with improved reliability and judgment.

## Overview

- **Provider:** Anthropic
- **Model ID:** `claude-opus-4-8`
- **Type:** Chat
- **Canonical URL:** https://modelstack.cc/models/chat/claude-opus-4-8
- **Version:** Opus 4.8
- **Knowledge cutoff:** 2026-01
- **Best for:** Complex multi-step agentic tasks, long-horizon coding, professional work, error detection, honest responses

## Pricing

| Usage | Price |
| --- | ---: |
| Input | $5 / 1M tokens |
| Output | $25 / 1M tokens |
| Cache read | $0.5 / 1M tokens |
| Cache write (5 minutes) | $6.25 / 1M tokens |
| Cache write (1 hour) | $10 / 1M tokens |

## Specifications

- **Context window:** 1M
- **Maximum output:** 128K
- **Typical latency:** 20 ms
- **Released:** 2026-05-28

## Capabilities

- chat
- tool_use
- reasoning
- vision
- coding
- analysis
- long_context

## API usage

Use this model ID in a ModelStack API request:

```json
{
  "model": "claude-opus-4-8"
}
```

For the rendered page, visit https://modelstack.cc/models/chat/claude-opus-4-8.