Overview
Claude Opus 4.8 is Anthropic's most capable Opus-tier model, positioned for complex reasoning, long-horizon agentic coding, and high-autonomy work. It became generally available on May 28, 2026 via the Claude API, Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry.
It ships with a 1M-token context window at standard pricing and supports up to 128K output tokens. The model accepts text and image input and uses adaptive thinking (always on); on Opus 4.8 the effort parameter defaults to high across the API and Claude Code.
Pricing is $5 per million input tokens and $25 per million output tokens, with cache reads at $0.50 per million. An optional research-preview Fast mode trades higher per-token cost ($10 input / $50 output) for significantly faster output.
| Released | 2026-05-28 |
|---|---|
| License | Proprietary |
| Weights | API only |
| Parameters | Undisclosed |
| Context | 1M |
| Max output | 128K |
| Knowledge cutoff | Jan 2026 |
| Modalities | Text, Vision |
| Status | Generally available |
Benchmarks
Claude Opus 4.8 benchmark comparison as published by Anthropic on the Claude Opus 4.8 launch page.
| Benchmark | Claude Opus 4.8 | Claude Opus 4.7 | GPT-5.5 | Gemini 3.1 Pro |
|---|---|---|---|---|
| Agentic coding (SWE-Bench Pro) | 69.2% | 64.3% | 58.6% | 54.2% |
| Agentic terminal coding (Terminal-Bench 2.1) | 74.6% | 66.1% | 78.2% | 70.3% |
| Multidisciplinary reasoning (Humanity's Last Exam, no tools) | 49.8% | 46.9% | 41.4% | 44.4% |
| Multidisciplinary reasoning (Humanity's Last Exam, with tools) | 57.9% | 54.7% | 52.2% | 51.4% |
| Agentic computer use (OSWorld-Verified) | 83.4% | 82.8% | 78.7% | 76.2% |
| Knowledge work (GDPval-AA) | 1890 score | 1753 score | 1769 score | 1314 score |
| Agentic financial analysis (Finance Agent v2) | 53.9% | 51.5% | 51.8% | 43% |
This model's scores
Scores on a 0–100 scale (25-point gridlines); higher is better. Each benchmark links to its published source.
Pricing
| Input | $5.00 / 1M tokens |
|---|---|
| Cached input | $0.50 / 1M tokens |
| Output | $25.00 / 1M tokens |
Strengths
- Frontier reasoning and long-horizon agentic coding
- 1M-token context window billed at standard per-token rates
- Adaptive thinking with a high default effort level for deep reasoning
- Strong agentic web/computer-use performance (84% on Online-Mind2Web)
- Direct reasoning over PDFs, diagrams, and other unstructured visual content
Best for
- Reach for it when you need the most capable model for multi-step agentic coding across a large codebase.
- Reach for it when a task needs deep reasoning over very long documents within a 1M-token context.
- Reach for it for high-autonomy agent workflows where reliability matters more than per-token cost.
How to access
| Provider | Model ID |
|---|---|
| Anthropic API ↗ | claude-opus-4-8 |
Claude Opus — every version
The full lineage of the Claude Opus line, newest first. Every version has its own page — click any to compare specs, benchmarks and pricing.
| Version | Released | Context | License |
|---|---|---|---|
| Claude Opus 4.8current | 2026-05-28 | 1M | Proprietary |
| Claude Opus 4.7 | 2026-04-16 | — | Proprietary |
| Claude Opus 4.6 | 2026-02-05 | — | Proprietary |
| Claude Opus 4.5 | 2025-11-24 | — | Proprietary |
| Claude Opus 4.1 | 2025-08-05 | — | Proprietary |
| Claude Opus 4 | 2025-05-22 | — | Proprietary |
| Claude 3 Opus | 2024-03-04 | — | Proprietary |
FAQ
How much does Claude Opus 4.8 cost?
Claude Opus 4.8 costs $5 per million input tokens and $25 per million output tokens on the Anthropic API. Prompt-cache reads are $0.50 per million tokens. An optional Fast mode raises this to $10 input and $50 output per million tokens for faster output.
What is the context window of Claude Opus 4.8?
Claude Opus 4.8 has a 1M-token context window, available at standard per-token pricing, and can generate up to 128K output tokens in a single synchronous response. On Microsoft Foundry the context window is 200K tokens.
When was Claude Opus 4.8 released?
Anthropic made Claude Opus 4.8 generally available on May 28, 2026, across the Claude API, Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry. Its reliable knowledge cutoff is January 2026.