Anthropic Claude Opus 4.7

Opus 4.7 is Anthropic's most capable generally available model, released on April 16, 2026. It delivers a 13% lift on coding benchmarks, 3x more production tasks resolved, and near-perfect vision accuracy at 98.5% — all at the same pricing as Opus 4.6. The model is purpose-built for long-horizon agentic work, complex software engineering, and sustained knowledge tasks that require deep reasoning and self-verification.

Beyond coding, Opus 4.7 introduces high-resolution vision support up to 3.75 megapixels (3x the previous limit), a new xhigh effort level for finer quality-cost control, and stricter instruction following that makes it more predictable in production. It scores 64.3% on SWE-bench Pro (up from 53.4%), 70% on CursorBench (up from 58%), and leads on agentic benchmarks ahead of GPT-5.4 and Gemini 3.1 Pro.

---

Why It Looks Great

• Large Language Model architecture with adaptive thinking for efficient processing
• 1000000 context window for long document handling and multi-session memory
• Competitive pricing at $5.0/$25.0 per million tokens (same as Opus 4.6)
• 98.5% visual acuity with 3.75MP image support — a transformational leap from 54.5%
• 3x more production tasks solved compared to its predecessor

---

Key Features

• Context Window: 1000000 tokens
• Max Output: 128000 tokens
• Vision: Supported (high-resolution up to 3.75 megapixels)
• Function Calling: Supported
• Effort Levels: low, medium, high, xhigh (new), max
• Task Budgets: Supported (public beta)

Benchmarks

Specifications

---

Pricing

Note: Opus 4.7 ships with an updated tokenizer that may produce 1.0–1.35x more tokens depending on content type. The per-token price is unchanged, but effective cost may increase up to 35% for certain content (structured data, code). Benchmark your actual workloads before migrating at scale.

---

How to Use

1. Write your prompt — describe the task, provide context, and specify desired output format.
2. Submit — the model processes your request and returns the response.

API Integration

Base URL: https://llm.wavespeed.ai/v1 API Endpoint: chat/completions Model ID: anthropic/claude-opus-4.7

---

API Usage

Python SDK

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://llm.wavespeed.ai/v1"
)

response = client.chat.completions.create(
    model="anthropic/claude-opus-4.7",
    messages=[{"role": "user", "content": "Hello!"}]
)

print(response.choices[0].message.content)

cURL

curl https://llm.wavespeed.ai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "model": "anthropic/claude-opus-4.7",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

---

What's New vs Opus 4.6

---

Notes

• Model: anthropic/claude-opus-4.7
• Provider: Anthropic
• Also available on: Amazon Bedrock, Google Cloud Vertex AI, Microsoft Foundry
• The model follows instructions more literally than Opus 4.6 — prompts may need retuning
• Extended thinking now uses adaptive mode only (manual budget_tokens removed)
• Sampling parameters (temperature, top_p, top_k) are no longer supported via API

Sources: Anthropic official release, felloai.com, nxcode.io, thenextweb.com. Content was rephrased for compliance with licensing restrictions.