# Claude Opus 5

> Claude Opus 5 is Anthropic's current flagship model, built for work that runs long: multi-file refactors, agentic coding loops, and analysis that has to hold a large codebase or document set in view at once. It takes a 1M-token context window and vision input, and it is the model Anthropic positions for tasks where giving up halfway is worse than being slow.

Canonical: https://routemux.com/anthropic/claude-opus-5
Provider: Anthropic

## Key facts

| Field | Value |
| --- | --- |
| Model ID | anthropic/claude-opus-5 |
| Official model ID | claude-opus-5 |
| Context window | 1M |
| Max output | 8K |
| Input modalities | text, image |
| Output modalities | text |
| Capabilities | Vision, Reasoning, Tool Calling, Coding, Cache, Strip Sampling Params |
| Upstream release | 2026-07-24 |
| Availability | available |

## Pricing

- RouteMux input: $1.4 per 1M tokens
- RouteMux output: $7 per 1M tokens
- Official input: $5 per 1M tokens
- Official output: $25 per 1M tokens
- Cache read: $0.14 per 1M tokens
- Cache write: $1.75 per 1M tokens
- Saving vs official: 72%
- Official price source: anthropic_x0.28, verified 2026-08-16

Billing is prepaid wallet, charged per successful request only.

## API access

Base URL: https://api.routemux.com

Supported protocols:

- anthropic_messages

## What is Claude Opus 5

Claude Opus 5 is Anthropic's current flagship model, built for work that runs long: multi-file refactors, agentic coding loops, and analysis that has to hold a large codebase or document set in view at once. It takes a 1M-token context window and vision input, and it is the model Anthropic positions for tasks where giving up halfway is worse than being slow.

Through RouteMux it costs $1.40 per 1M input tokens against Anthropic's $5, and $7 per 1M output against $25 — the same model and the same Anthropic Messages API, routed through one key alongside every other model in the catalog. Failed and timed-out requests are not billed, and prompt caching is priced separately at $0.14 read / $1.75 write per 1M tokens, which matters more on Opus than on cheaper models because the base rate is what caching is discounting.

## Key capabilities

- **1M-token context** — Enough to hold a mid-sized repository or a full document set in a single request, so long-horizon tasks do not need retrieval plumbing to get started.
- **Vision input** — Accepts images alongside text — screenshots, diagrams, and scanned documents go in the same message as the prompt.
- **Tool calling** — Native function calling, which is what makes the agentic loops it is built for possible without prompt-level hacks.
- **Prompt caching** — Cache reads are $0.14 / 1M against a $1.40 base rate. On a long system prompt reused across a session this is where most of the saving actually comes from.

## Other variants in this family

- Claude Opus 4.7 (`anthropic/claude-opus-4.7`)
- Claude Opus 4.6 (`anthropic/claude-opus-4.6`)
- Claude Opus 4.8 (`anthropic/claude-opus-4.8`)

## Sources

- [Anthropic — Models overview](https://docs.anthropic.com/en/docs/about-claude/models)
- [Anthropic — Pricing](https://www.anthropic.com/pricing)

_Content reviewed by Jerry Fan on 2026-08-09._