Enterprise DNAEnterprise DNA
Anthropic logo

Models · Best coding

Claude Opus 4

by Anthropic

Claude Opus 4 is benchmarked as the world’s best coding model, at time of release, bringing sustained performance on complex, long-running tasks and agent workflows.

Visit Anthropic ↗

Our verdict

Not reviewed yet

We have not reviewed Claude Opus 4 yet

No one at Enterprise DNA has checked this listing against its sources. The description comes from public listings gathered automatically. Any fact below shows where it came from and when it was last checked.

How we review records

From the listing

About Claude Opus 4

Claude Opus 4 is benchmarked as the world’s best coding model, at time of release, bringing sustained performance on complex, long-running tasks and agent workflows.

Best for

Hard reasoning, architecture reviews, and high-stakes agent plans

Use cases

  • architecture reviews
  • complex debugging
  • strategy memos
  • eval judges

Enterprise DNA angle

Use Opus where judgment matters; Omni routes cheaper models for bulk work and reserves Opus for the hard steps.

Notes

Claude Opus 4 is listed in the EDNA Models directory from the OpenRouter catalogue.

Pros

  • Top-tier reasoning for hard planning and analysis
  • Strong coding on complex multi-step tasks
  • Vision support for screenshots and docs

Cons

  • Higher $/Mtoken than Sonnet or open models
  • Overkill for simple classification or extraction

Key facts

What we have checked

Gpqa diamond
76.3% (for 22 May 2025)
Checked Epoch AI (CC BY 4.0)
Swe bench verified
70.7% (for 22 May 2025)
Checked Epoch AI (CC BY 4.0)
Arc agi 2
8.6% (for 22 May 2025)
Checked Epoch AI (CC BY 4.0)
Aider polyglot
72% (for 22 May 2025)
Checked Epoch AI (CC BY 4.0)
Arc agi
35.7% (for 22 May 2025)
Checked Epoch AI (CC BY 4.0)
Cybench
38% (for 22 May 2025)
Checked Epoch AI (CC BY 4.0)
Deepresearch bench
46.8% (for 22 May 2025)
Checked Epoch AI (CC BY 4.0)
Dtbench
81.6% (for 22 May 2025)
Checked Epoch AI (CC BY 4.0)
Fiction livebench
61.1% (for 22 May 2025)
Checked Epoch AI (CC BY 4.0)
Frontiermath 2025 02 28 private
4.5% (for 22 May 2025)
Checked Epoch AI (CC BY 4.0)
Frontiermath tier 4 2025 07 01 private
4.2% (for 22 May 2025)
Checked Epoch AI (CC BY 4.0)
Geobench
49% (for 22 May 2025)
Checked Epoch AI (CC BY 4.0)

Each value shows when we last confirmed it and where it came from. Values we have not confirmed are left out, never guessed.

API access

How to call it

Call via OpenRouter with model id `anthropic/claude-opus-4`.

  • openrouter · anthropic/claude-opus-4 · Docs

OpenRouter id: anthropic/claude-opus-4

Scores

How Claude Opus 4 scores

Each score comes from the named source and shows when we last confirmed it. A flag means the score needs care.

ScoreValueSourceChecked
Aider polyglot (best-across-versions)72percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Arc agi (best-across-versions)35.7percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Arc agi 2 (best-across-versions)8.6percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Cybench (best-across-versions)38percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Deepresearch bench (best-across-versions)46.8percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Dtbench (best-across-versions)81.6percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Fiction livebench (best-across-versions)61.1percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Frontiermath 2025 02 28 private (best-across-versions)4.5percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Frontiermath tier 4 2025 07 01 private (best-across-versions)4.2percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Geobench (best-across-versions)49percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Gpqa diamond (best-across-versions)76.3percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Gso bench (best-across-versions)6.9percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Show 9 more scores
ScoreValueSourceChecked
Lech mazur writing (best-across-versions)83.6percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Lmca (best-across-versions)37.4percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Math level 5 (best-across-versions)85percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Metr time horizons (best-across-versions)63.9percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Otis mock aime 2024 2025 (best-across-versions)Saturated64.4percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Simplebench (best-across-versions)58.8percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Swe bench verified (best-across-versions)70.7percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Vpct (best-across-versions)38percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Weirdml (best-across-versions)43.7percentAs of 22 May 2025Epoch AI (CC BY 4.0)
Saturated
Top models score close to the maximum, so this test no longer separates them.

Sources

Where this comes from

A source shows what was published when we checked it. It does not certify security, performance or a working installation. Read our method.

Changelog

What changed

No changes recorded yet. Releases, price changes and deprecations appear here, with their dates, once we track them for this record.

MODEL FIELD GUIDE / SOURCE EVIDENCE

What you can do with Claude Opus 4.

No longer listed by its catalogue; the facts shown are as last listed. Sources checked 2026-11-07. Hosting routes and app features can differ.

DeveloperAnthropic
Exact model identifieranthropic/claude-opus-4
InputNot established
OutputNot established
Tool calling
Not established
Structured output
Not established
Reasoning controls
Not established
Availability
unknown
Price basis
token
Licence verification
Check the developer licence; the model name is insufficient
Context length
Not established
Knowledge cutoff
Not established
Developer release date
Not established
First seen in the catalogue
Not established
Listed retirement date
Not established
Last listed by its catalogue
20 Sept 2026

Not in the OpenRouter catalogue when checked; last listed 2026-09-20. Retirement is not confirmed by a source.

Inspect the sources and their scope
How to evaluate a model for your work →

FROM SPECIFICATIONS TO YOUR WORK

Is this the right model for your job?

Start with a realistic sample. These workflows match the model’s listed capabilities, not a verified result on your task.

Write a clear business documentWorkflow & starter prompt ↗Analyse a spreadsheetWorkflow & starter prompt ↗Research a companyWorkflow & starter prompt ↗Build an appWorkflow & starter prompt ↗

What would it cost?

Pick a common job and how many times a month you run it. Token sizes are our assumption for a typical run, shown under the result.

Working out the cost...

Token prices only. Per-request, image, tool, storage and subscription fees are not included. One token is about three quarters of an English word. Real counts vary by model and language. How tokens are counted ↗

Check the source

Price date: 20 Sept 2026, from the price listed on this record. Prices are USD API token rates. Your app subscription, uploads, browsing and tools are separate.

Model source & availability ↗

Understand the evidence

Listed capabilities and third-party benchmark scores do not establish Enterprise DNA task-test results. Validate accuracy and total cost on your own sample.

Compare this model with alternatives →
Free 27-page guide

Get the free Developer’s Field Guide

A 27-page field guide to the AI coding workflow with Claude. Claude Code, MCP servers, the prompt patterns that work, and what to delegate. Free.

Enter your work email. We send it straight over, plus a few short notes worth knowing. Unsubscribe any time.

No spam. Unsubscribe any time.

← Back to ModelsSubmit your own entry →