Anthropic · Released May 22, 2025Claude Opus 4
Flagship Claude model for deep reasoning, coding, and long-horizon agents.
- Capability
- 142.7ECI · #87 of 148
- Input
- $15.00per 1M tokens
- Output
- $75.00per 1M tokens
- Context
- 200K32K max output
What it costs
- Input
- $15.00per million tokens
- Output
- $75.00per million tokens
- Blended (3:1)
- $30.00pricier than 97% of priced models
Typical monthly bills
| Monthly usage | Estimated cost |
|---|---|
| Side project2M input + 0.5M output tokens | $67.50 |
| Team assistant25M input + 5M output tokens | $750.00 |
| Production app250M input + 50M output tokens | $7,500 |
Median of 5 paid API providers on models.dev.
Independent benchmarks
#87 of 148 scored models
The shaded band is Epoch AI’s confidence range (140.2–144.2); the tick marks the median scored model.
Scores from Epoch AI, run independently of Anthropic.
The details
- Lab
- Anthropic
- Released
- May 22, 2025
- Knowledge cutoff
- Mar 31, 2025
- Context window
- 200,000 tokens
- Max output
- 32,000 tokens
- Inputs
- Text, Images, PDFs
- Output
- Text
- Reasoning
- Always on
- Tool calling
- Yes
- Structured output
- No
- Weights
- Proprietary
- Availability
- 5 API providerslisted on models.dev
What Anthropic claims
Published by the lab at launch. Settings vary, so compare these only with care.
| Benchmark | Score | Setting | Source |
|---|---|---|---|
| Aider Polyglot | 72 percent correct | — | Source |
More from Anthropic
About Claude Opus 4
How much does Claude Opus 4 cost?
Claude Opus 4 costs $15.00 input / $75.00 output per million tokens (median across 5 API providers). At a 3:1 input-to-output mix that is $30.00 per million tokens, more expensive than 97% of the 360 priced models we track.
What is the context window of Claude Opus 4?
Claude Opus 4 accepts up to 200,000 tokens per request and can write up to 32,000 tokens in one response.
How good is Claude Opus 4?
Epoch AI gives Claude Opus 4 a Capabilities Index score of 142.7 (likely range 140.2–144.2), ranking it #87 of 148 models Epoch has scored. Epoch AI benchmark results: GPQA Diamond 76.3%, OTIS Mock AIME 2024–2025 64.4%, SWE-bench Verified 70.7%.
Is Claude Opus 4 open source?
No. Claude Opus 4 is proprietary; you use it through an API or partner platforms.
What inputs does Claude Opus 4 support?
Claude Opus 4 accepts text, images and PDFs and replies in text. It is a reasoning model, supports tool calling.
When was Claude Opus 4 released?
Anthropic released Claude Opus 4 on May 22, 2025. Its training data runs to Mar 31, 2025.
What are the best alternatives to Claude Opus 4?
The closest current models from other labs on capability, price and release date are o3-pro (OpenAI, ECI 147.4, $20.00 / $80.00), DeepSeek-R1-Distill-Qwen-32B (DeepSeek, ECI 137.4), Gemini 2.5 Pro (Google, ECI 145.3, $1.25 / $10.00) and Qwen3 Max (Alibaba (Qwen), ECI 142.4, $1.20 / $6.00).