Skip to content
VECTOR WIREAI INTELLIGENCE
⌘K
UTC
Models741Deals30d270Policy30d65, 13 pendingSources1,015

Claude Opus 5

AnthropicReleased Jul 24, 2026Current

Score basis

Claude Opus 5 is at the frontier in agentic tasks, reasoning, and coding; strong in factuality; and capable in math, long context, and multimodal tasks. Too few results yet to rate safety, multilingual tasks, or instruction following.

Capability profile

Bars show the model's median result as a share of the leading model's, per capability. Select a row to see its results.

Frontier3 capabilities
  1. AgenticLeads7 of 7
  2. Reasoning−2.3%6 of 6
  3. Coding−5.9%8 of 10
Strong1 capability
  1. Factuality−8.7%3 of 4
Capable3 capabilities
  1. Math−12.3%5 of 5
  2. Long Context−12.5%1 of 3
  3. Multimodal−18.6%2 of 6
Not rated3 capabilities

Too few results yet to rate Safety, Multilingual or Instruction Following.

Price

$5.00input$25.00outputper million tokens

From Anthropic's own price page · 4 providers tracked · All prices

Costs more than 93% of 330 priced models · 3:1 input-to-output blend, log scale

Evidence

174results on78benchmarks

  • 22 independently verified
  • 79 aggregator
  • 57 vendor-reported
  • 16 cross-referenced

From 31 sources · latest Oct 8, 2026 · How verification works

API features

Tool callingStructured outputsReasoning

As listed by OpenRouter

Research

65 papers reference Claude Opus 5

Claude Opus 5 benchmark results

174 results on 78 benchmarks, grouped by capability. Each bar is the result as a share of the capability leader's; every result links to its source.

Agentic

FrontierFull agentic ranking

Leads the field7 of 7 ranked benchmarks measured

Show 13 more agentic resultsHide 13 agentic results

Reasoning

FrontierFull reasoning ranking

2.3% behind the leader6 of 6 ranked benchmarks measured

Show 21 more reasoning resultsHide 21 reasoning results

Coding

FrontierFull coding ranking

5.9% behind the leader8 of 10 ranked benchmarks measured

Show 10 more coding resultsHide 10 coding results

Factuality

StrongFull factuality ranking

8.7% behind the leader3 of 4 ranked benchmarks measured

Show 8 more factuality resultsHide 8 factuality results

Math

CapableFull math ranking

12.3% behind the leader5 of 5 ranked benchmarks measured

Long Context

CapableFull long context ranking

12.5% behind the leader1 of 3 ranked benchmarks measured

Show 4 more long context resultsHide 4 long context results

Multimodal

CapableFull multimodal ranking

18.6% behind the leader2 of 6 ranked benchmarks measured

Show 5 more multimodal resultsHide 5 multimodal results

Instruction Following

Not enough dataFull instruction following ranking

0 of 3 ranked benchmarks measured

More results

Benchmarks outside the capability baskets. They are not ranked against a leader.

Show 77 more resultsHide 77 results

Claude Opus 5: common questions

Who makes Claude Opus 5?

Claude Opus 5 is made by Anthropic.

When was Claude Opus 5 released?

Claude Opus 5 was released on Jul 24, 2026, according to Anthropic's own announcement.

What is Claude Opus 5 good at?

Claude Opus 5 is at the frontier in agentic tasks, reasoning, and coding; strong in factuality; and capable in math, long context, and multimodal tasks. Too few results yet to rate safety, multilingual tasks, or instruction following.

How much does Claude Opus 5 cost?

Claude Opus 5 costs $5.00 per million input tokens and $25.00 per million output tokens, according to Anthropic's own price page. We track its price at 4 providers. At a mix of three input tokens to one output token, it costs more than 93% of the 330 priced models we track.

How many benchmarks has Claude Opus 5 been tested on?

We track 174 results for Claude Opus 5 on 78 benchmarks from 31 sources, 22 of them independently verified. The latest was recorded on Oct 8, 2026.

Which API features does Claude Opus 5 support?

OpenRouter lists tool calling, structured outputs, and reasoning for Claude Opus 5.

About this record

Where Claude Opus 5's numbers come from, and every name it appears under.

Tracked since
Jul 17, 2026
Newest source mention
Oct 5, 2026

Where the results come from

Verification: 174 scores · 22 independently verified · 79 aggregator-attributed · 16 vendor cross-reference · 57 vendor-reported. How these tiers are assigned

From 31 sources on 13 sites. Artificial Analysis supplies 80 of them; the 22 independently verified results come from 8 sites. Bars are coloured by trust tier.

  • artificialanalysis.ai80
  • www-cdn.anthropic.com37
  • anthropic.com12
  • huggingface.co9
  • api.llm-stats.com8
  • deepmind.google7
  • livebench.ai7
  • arcprize.org4
  • datasets-server.huggingface.co3
  • epoch.ai3
  • labs.scale.com2
  • lmarena.ai1
  • simple-bench.com1

Also known as

How our sources name Claude Opus 5 at each reasoning setting.

SettingShort formLong formAPI id
lowclaude opus 5 (low)claude opus 5 (adaptive reasoning, low effort)claude-opus-5-low
mediumclaude opus 5 (medium)claude opus 5 (adaptive reasoning, medium effort)claude-opus-5-medium
highclaude opus 5 (high)claude opus 5 (adaptive reasoning, high effort)claude-opus-5-high
xhigh—claude opus 5 (adaptive reasoning, xhigh effort)claude-opus-5-xhigh claude-opus-5 (xhigh)
maxclaude opus 5 (max)claude opus 5 (adaptive reasoning, max effort)claude-opus-5-max
Also listed asopus 5claude opus5opus5