Anthropic: Claude Opus 4.7

Stronger Opus tier for advanced software work and high-stakes reasoning

claude-opus-4-7

AvailableAnthropicChatCached input 90% less

Input / Output-70%: $5.00 / $25.00$1.50 / $7.50
Context: 1M
Released: Apr 16, 2026
Max output: 128K
Modalities: VisionChat
Capabilities: Tool usePrompt CacheReasoning

About Claude Opus 4.7

Claude Opus 4.7 is Anthropic's Opus-tier model from April 2026 for advanced software engineering and high-stakes reasoning. Compared with Opus 4.6 it reads much larger images, follows instructions more literally, adds an xhigh effort level and task budgets, and ships a new tokenizer that later Claude models also use. Anthropic now lists it as a legacy model.

Where it works well

Anthropic reports a clear gain over Opus 4.6 on CursorBench, a gain aimed at long, hard coding tasks.
Accepts images up to 2,576 pixels on the long edge, so computer-use agents can read dense screenshots and complex diagrams.
Follows instructions more literally than Opus 4.6, which makes constrained output formats and strict checklists easier to enforce.
The xhigh effort level sits between high and max, giving finer control over the reasoning versus latency trade.

When to choose another model

Literal instruction following means prompts written for looser models can produce surprising results until you retune them.
The new tokenizer maps the same text to roughly 1.0 to 1.35 times more tokens, which affects length budgeting.
Anthropic says it is less capable than Claude Mythos Preview, and it is a legacy model next to Opus 5.5.

Getting started

Create API key
Create a key in Console, then use it with every model on the platform.

Send your first request

Copy the example for your language and run it against the endpoint.

Text to textMessages API

POST/v1/messages

API format:

curl https://api.tokenlab.sh/v1/messages \
  -H "Content-Type: application/json" \
  -H "x-api-key: sk-xxx" \
  -H "anthropic-version: 2023-06-01" \
  -d '{
    "model": "claude-opus-4-7",
    "max_tokens": 1024,
    "messages": [
      {"role": "user", "content": "Hello!"}
    ]
  }'

Pricing

The Verified price applies to Verified, which costs less on most models. The Official price is the model maker's published price and applies to the more reliable Official route. Auto bills the route that completes the request.

Default spec

per 1M tokens

Official price: Input $5.00 / Output $25.00 / Cache read $0.50 / Cache write $6.25 / Cache write (1h) $10.00
Verified price: Input $1.50 / Output $7.50 / Cache read $0.15 / Cache write $1.88 / Cache write (1h) $3.00
Discount: -70%

Prompt cache pricing

Cache read

Official price: $0.50
Verified price: $0.15
Discount: -70%

Cache write

Official price: $6.25
Verified price: $1.88
Discount: -70%

Cache write (1h)

Official price: $10.00
Verified price: $3.00
Discount: -70%

Tool fees

Charged per call, only when the model uses the tool.

Web search

Official price: $0.01/search
Verified price: $0.003/search
Discount: -70%

	Official priceper 1M tokens	Verified priceper 1M tokens	Discount
Default spec	Input $5.00 / Output $25.00 / Cache read $0.50 / Cache write $6.25 / Cache write (1h) $10.00	Input $1.50 / Output $7.50 / Cache read $0.15 / Cache write $1.88 / Cache write (1h) $3.00	-70%
Prompt cache pricing
Cache read	$0.50	$0.15	-70%
Cache write	$6.25	$1.88	-70%
Cache write (1h)	$10.00	$3.00	-70%
Tool feesCharged per call, only when the model uses the tool.
Web search	$0.01/search	$0.003/search	-70%

Usage & activity

Success rate is the share of requests that completed. Latency is how long a full response takes; P95 means 95% of requests finished within that time.

Usage & availability

Last 24 hours

Requests
Success rate
P95 latency
Total tokens

Model performance

Last updated Nov 1, 2026, 8:13 AM UTC

Metrics appear once privacy and data volume thresholds are met.

Data is based on aggregate user requests, excluding status checks.

Open in Console

Open Claude Opus 4.7 in Console with a prompt ready to edit or send.

Help me try claude-opus-4-7 with a short message at /v1/messages. Show the reply, latency, and cost.

Use cases

Best for

Reasoning
Vision

Hand-off coding tasks
Give it a spec and a repository and let it work through a feature across files, using xhigh effort when the first pass at high misses edge cases.
Screenshot and diagram reading
Feed full-resolution dashboards, architecture diagrams, or scanned forms and have it extract values without downscaling the image first.
Document reasoning
Review contracts, financial statements, or lab reports where Anthropic reports gains over Opus 4.6 in finance, legal, and document evaluations.

Prompt examples

Implement the feature described in SPEC.md, update the tests, and list any requirement you were unable to satisfy.

Read this architecture diagram and write out every service-to-service call as a table with protocol and direction.

Compare these two vendor agreements clause by clause and flag differences in liability and termination.

This model has conditional pricing. Monthly token totals alone cannot produce a reliable estimate; use the detailed pricing for the request specification, cache, and applicable time window.

FAQ

What changed from Claude Opus 4.6 to Opus 4.7?

Opus 4.7 follows instructions more literally, accepts much larger images, adds the xhigh effort level and task budgets in beta, and uses a new tokenizer. Anthropic reports higher scores than 4.6 on coding, finance, legal, document reasoning, and biology evaluations.

Does Claude Opus 4.7 see images in more detail?

Yes. Anthropic states it accepts images up to 2,576 pixels on the long edge, about 3.75 megapixels and more than three times what earlier Claude models took. That helps with dense screenshots, small chart labels, and technical diagrams.

Will my Opus 4.6 prompts work on Opus 4.7?

Mostly, but retest them. Anthropic warns that 4.7 interprets directives more literally, so a prompt that relied on the model guessing intent can now do exactly what it says. Also expect different token counts from the new tokenizer.

Is Claude Opus 4.7 the latest Opus model?

No. Anthropic lists Opus 4.7 as legacy and recommends Opus 5.5 for current work. Opus 4.7 remains useful when a pipeline is validated on it, or when you want its behavior fixed.

How much does Claude Opus 4.7 cost?

On TokenLab, Claude Opus 4.7 costs Input $1.50 / Output $7.50 per 1M tokens. The pricing table above shows the full breakdown. Rates depend on the billing unit, specification, and usage. Compare matching conditions in the model's detailed pricing; a single rate does not determine the total cost.

What are the context window and output limit of Claude Opus 4.7?

Claude Opus 4.7 accepts up to 1,000,000 tokens of context and returns up to 128,000 tokens in one response.

Which endpoint should Claude Opus 4.7 use?

Use https://api.tokenlab.sh/v1/messages for Claude Opus 4.7. The request example below shows the matching code shape.

Which operations does Claude Opus 4.7 support?

Claude Opus 4.7 supports Text to text. Select an operation above to see its endpoint and request example.