Claude Opus 5 (batch)
anthropic/claude-opus-5:batchAnthropic’s reasoning model for demanding software, coding, visual analysis, and long multi-step tasks.
1M tokens of context — about 750,000 words, on a scale that ends at 2M.
What it can do
- Reasoning
- Vision
- Tool use
- Takes text
- Takes image
- Takes file
- Gives text
Released 24 Jul 2026 · checked 3 Oct 2026.
About it
Claude Opus 5 (batch) is an Anthropic model released July 24, 2026. It accepts text, images, and files, can use tools, and returns text. Its one-million-token context lets it take in very large inputs, while its maximum answer is 128,000 tokens. It is aimed at demanding reasoning, coding, code review, bug finding, visual analysis, and extended multi-step work.
Review
In a line
Claude Opus 5 (batch) is Anthropic’s model for difficult reasoning, software work, and extended multi-step tasks. Its model ID is anthropic/claude-opus-5:batch.
What it's good at
It suits demanding reasoning, coding, code review, and finding bugs. It is also designed for work that runs through many steps from start to finish.
The model can analyze images as well as text and files. It supports tool use, which lets a connected application give it actions to perform.
Price in plain terms
Input costs $2.5 per million tokens. Output costs $12.5 per million tokens. Tokens are small pieces of text or other input that the model reads or produces.
At the listed rates, reading a roughly 500-word page costs about 0.16 cents. A chat turn with about 1,000 input tokens and 500 output tokens costs about 0.88 cents; 1,000 such turns cost about $8.75.
Context and limits
Claude Opus 5 can take up to 1,000,000 tokens in one request, so it can handle very large amounts of material at once. It can produce up to 128,000 tokens in one answer.
It accepts text, images, and files, but its output is text. It supports reasoning, image understanding, and tool use.
Ways to use it
Use the model in a service or application that offers Claude Opus 5 (batch). Applications can send text, images, or files, and can connect tools for the model to use.
For model directories or API settings, the identifier is anthropic/claude-opus-5:batch.
Verdict
Choose this model when a task needs careful reasoning, software coding or review, image understanding, and a very large working space. Its high output price matters most when answers are long or used often.
It is a poor fit if you only need short text replies, do not need images or tools, or must keep token spending as low as possible.
Verdict
Claude Opus 5 (batch) makes sense for teams handling complex software work, large documents, or tasks that combine text, images, files, and connected tools. The million-token context is useful when a job cannot be split into small pieces. Its output rate is much higher than its input rate, so frequent, lengthy answers can add up. Pick a simpler or cheaper model when your requests are brief and routine, and avoid relying on it for non-text output because it returns text only.
- Handles up to 1,000,000 input tokens at once
- Supports text, images, and files as inputs
- Reasoning, tool use, and vision are supported
- Up to 128,000 tokens in one answer
- Strong fit for coding, code review, and bug finding
- Output costs $12.5 per million tokens
- Returns text only
- Pick it ifDemanding coding, code review, and bug finding; Large, multi-step tasks using text, images, or files; Applications that need reasoning and tool use
- Skip it ifYou need non-text output; Your work is short and routine; You need a listed free tier or built-in integrations
Pricing
Input is $2.5 per million tokens and output is $12.5 per million tokens. A 500-word page costs about 0.16 cents to read; a typical chat turn costs about 0.88 cents, or about $8.75 for 1,000 turns.
Questions
Who makes Claude Opus 5 (batch)?
Anthropic makes it.
What can I send to the model?
It accepts text, images, and files.
Can it use tools?
Yes. The model supports tool use when an application connects tools to it.
How much can it read at once?
Its context window is 1,000,000 tokens.
How long can one answer be?
The maximum output is 128,000 tokens.
What does it cost?
Input is $2.5 per million tokens and output is $12.5 per million tokens.
What does a typical chat turn cost?
A turn with about 1,000 input tokens and 500 output tokens costs about 0.88 cents.
What is its model ID?
The model ID is anthropic/claude-opus-5:batch.
On these boards
More from Anthropic
- Claude Sonnet 5.5$2 in
- Claude Sonnet 5.5 (batch)$1 in
- Claude Opus 5.5$4 in
- Claude Opus 5.5 (batch)$2 in
- Claude Fable 5.1$10 in
- Claude Fable 5.1 (batch)$5 in
- Claude Opus 5$5 in
- Claude Sonnet 5$2 in
- Claude Sonnet 5 (batch)$1 in
- Claude Fable 5$10 in
- Claude Fable 5 (batch)$5 in
- Claude Fable Latest$10 in
- Claude Opus 4.8$5 in
- Claude Opus 4.8 (batch)$2.50 in
- Claude Haiku Latest$1 in
- Claude Sonnet Latest$2 in
- Claude Opus Latest$4 in
- Claude Opus 4.7$5 in