Anthropic/Model/Claude/4.8/Opus
Key Features
- Generally available as of May 28, 2026; Anthropic's most capable generally available model for complex reasoning, long-horizon agentic coding, and high-autonomy work
- Context window: 1M input tokens at standard API pricing
- Max output: 128k tokens on the synchronous Messages API (300k on Message Batches API with
output-300k-2026-03-24 beta header) - Adaptive thinking only: uses
thinking: {"type": "adaptive"}; extended thinking with explicit budget_tokens is not supported - Effort parameter defaults to
high on all surfaces including Claude API and Claude Code; set explicitly to use a different level - ~4x less likely than Opus 4.7 to overlook flaws in generated code
- Superior judgment in agentic tasks: better at catching mistakes and questioning unsound plans
- Enhanced tool calling efficiency
- Stronger performance on legal and financial workflows
- Improved multimodal reasoning over documents and diagrams
- Dynamic workflows for Claude Code: parallel subagent processing
- Messages API now accepts system entries mid-conversation
- Reliable knowledge cutoff: January 2026; training data cutoff: January 2026
- Model version status: Stable (general availability)
Performance Highlights
- 84% on Online-Mind2Web (computer-use benchmark); meaningful improvement over Opus 4.7 and GPT-5.5
- Highest-ever score on Legal Agent Benchmark; first model to exceed 10% on the all-pass standard
- Exceeds Opus 4.7 across all effort levels on CursorBench
- Outperforms GPT-5.5 on multiple proprietary evaluations including the Super-Agent benchmark
Model Tiers
- Flagship Opus release succeeding Anthropic/Model/Claude/4.7/Opus in the Claude 4 generation
- Earlier Opus pages in this garden:
- Speed-focused sibling: Claude Sonnet 4.6 (
claude-sonnet-4-6)
Access Details
- API availability: Claude Messages API model string
claude-opus-4-8 (dateless format, pinned snapshot per Anthropic versioning starting with the 4.6 generation) - Pricing (standard tier, per million tokens): $5 input · $25 output
- Fast mode pricing: $10 input · $50 output
- Batch API: 50% discount; prompt caching: up to 90% cost savings
- Platforms: Claude Pro, Max, Team, and Enterprise; Claude API; Amazon Bedrock; Google Cloud Vertex AI; Microsoft Foundry (note: 200k-token context window on Microsoft Foundry)
- Amazon Bedrock model ID:
anthropic.claude-opus-4-8 - Google Vertex AI ID:
claude-opus-4-8 - SDK support:
- Python:
pip install -U "anthropic" - TypeScript:
@anthropic-ai/sdk
Technical Specifications
- Model architecture: Proprietary Anthropic large language model (undisclosed weights)
- Input/output formats: Text and images in; text (and tool use) out
- Special capabilities:
- Adaptive thinking
- Dynamic workflows for parallel subagent processing in Claude Code
- Improved multimodal document and diagram reasoning
- Model version code:
claude-opus-4-8
References