Claude Opus 4.1
Claude Opus 4.1 improves on Claude Opus 4 with targeted gains in agentic task execution, real-world coding, and multi-step reasoning, available on AI Gateway via anthropic, bedrock, vertexAnthropic with automatic retries and failover.
import { streamText } from 'ai'
const result = streamText({ model: 'anthropic/claude-opus-4.1', prompt: 'Why is the sky blue?'})Playground
Try out Claude Opus 4.1 by Anthropic. Usage is billed to your team at API rates. Free users (those who haven't made a payment) get $5 of credits every 30 days.
Providers
Route requests across multiple providers. Copy a provider slug to set your preference. Visit the docs for more info. Using a provider means you agree to their terms, listed under Legal.
| Provider |
|---|
| 200K | 1.7s | 44tps | $15.00/M | $75.00/M | Read:$1.5/M Write: $18.75/M | $10.00/K + input costs | — | +3 | 05/22/2025 | |||
| 200K | 2.9s | 20tps | $15.00/M | $75.00/M | Read:$1.5/M Write: $18.75/M | — | — | +3 | 05/22/2025 | |||
| 200K | 1.3s | 42tps | $15.00/M | $75.00/M | Read:$1.5/M Write: $18.75/M | $10.00/K + input costs | — | +3 | 05/22/2025 |
P50 throughput on live AI Gateway traffic, in tokens per second (TPS). Visit the docs for more info.
P50 time to first token (TTFT) on live AI Gateway traffic, in milliseconds. View the docs for more info.
Direct request success rate on AI Gateway and per-provider. Visit the docs for more info.
More models by Anthropic
| Model |
|---|
| 1M | 6.8s | 96tps | $10.00/M | $50.00/M | Read:$1/M Write: $12.5/M | $10/K + input costs | — | +3 |  | 06/09/2026 | |||
| 1M | 2.2s | 119tps | $5.00/M | $25.00/M | Read:$0.5/M Write: $6.25/M | $10/K + input costs | — | +4 |   | 05/28/2026 | |||
| 1M | 0.8s | 94tps | $5.00/M | $25.00/M | Read:$0.5/M Write: $6.25/M | $10/K + input costs | — | +4 | | 04/16/2026 | |||
| 1M | 1.0s | 50tps | $3.00/M | $15.00/M | Read:$0.3/M Write: $3.75/M | $10/K + input costs | — | +4 | | 02/17/2026 | |||
| 1M | 0.8s | 59tps | $5.00/M | $25.00/M | Read:$0.5/M Write: $6.25/M | $10/K + input costs | — | +4 | | 02/05/2026 | |||
| 200K | 0.5s | 113tps | $1.00/M | $5.00/M | Read:$0.1/M Write: $1.25/M | $10.00/K + input costs | — | +3 | | 10/15/2025 |
About Claude Opus 4.1
Claude Opus 4.1 became available on AI Gateway on May 22, 2025, the same day Anthropic released the model. Access it via the AI SDK, Chat Completions API, Responses API, Messages API, or other API formats, from TypeScript or Python, with the model string anthropic/claude-opus-4.1. Through AI Gateway, you get a unified API, usage and cost tracking, and provider routing across anthropic, bedrock, vertexAnthropic, a single integration point without separate credentials for each provider.
Claude Opus 4.1 slots into existing Opus 4 workflows as an upgrade on the same task categories where Opus excels: complex multi-step agentic work, codebases requiring deep understanding across many files, and reasoning-intensive analysis where the model needs to maintain coherence over extended sessions.
The upgrade path from Opus 4 is a model string change. The API surface, pricing structure, and integration patterns stay consistent with the Opus 4 baseline.
What To Consider When Choosing a Provider
- Configuration: AI Gateway routes Claude Opus 4.1 across anthropic, bedrock, vertexAnthropic. Provider availability changes over time, so check the providers panel on this page when planning failover for critical production pipelines.
- Zero Data Retention: AI Gateway supports Zero Data Retention for this model via direct gateway requests (BYOK is not included). To configure this, check the documentation.
- Authentication: AI Gateway authenticates requests using an API key or OIDC token. You do not need to manage provider credentials directly.
When to Use Claude Opus 4.1
Best For
- Complex agentic task execution: Multi-step pipelines that require sustained planning and execution, the primary improvement area in this release
- Real-world coding tasks: Multi-file refactoring, debugging hard-to-reproduce issues, and understanding large unfamiliar codebases at Opus-tier depth
- Reasoning-intensive analysis: Depth of multi-step inference outweighs response speed
- Teams already on Opus 4: A targeted upgrade without changing your integration architecture
- Critical production pipelines: Workloads that require the Claude Opus 4.1 checkpoint specifically, with AI Gateway failover across anthropic, bedrock, vertexAnthropic
Consider Alternatives When
- Primary cost constraint: Sonnet-tier models deliver strong results at lower per-token cost
- Sonnet-sufficient tasks: A Sonnet variant fits task profiles that don't need Opus-depth reasoning
- Hard latency requirements: Opus models trade speed for depth
Conclusion
Claude Opus 4.1 is the right incremental upgrade for teams running Opus 4 in production who want improvements in agentic task execution, coding, and reasoning without a model architecture change. The upgrade path is a model string update, and AI Gateway handles the rest.
Frequently Asked Questions
What specific improvements does Claude Opus 4.1 make over Opus 4?
Claude Opus 4.1 improves agentic task execution, real-world coding, and reasoning over the Opus 4 baseline. It refines multi-step workflows and complex code generation while keeping the same API interface.
How do I switch from Claude Opus 4 to Claude Opus 4.1 in my existing code?
Update the model string to
anthropic/claude-opus-4.1in your AI SDK v5 integration. No other code changes are required for the common case.Which providers back Claude Opus 4.1 on AI Gateway?
AI Gateway routes Claude Opus 4.1 through anthropic, bedrock, vertexAnthropic, with automatic failover between providers.
Is Claude Opus 4.1 a different model from Opus 4, or an incremental update?
Claude Opus 4.1 is a targeted advance on the Opus 4 baseline, not a new architecture. Expect refinements in agentic task execution, coding, and reasoning rather than a fundamentally different model.
Does Claude Opus 4.1 support extended thinking with tool use?
Claude Opus 4.1 is part of the Claude 4 generation, which introduced extended thinking with tool use. Check the API documentation for the specific capabilities available on this checkpoint.
What latency should I expect from Claude Opus 4.1 through AI Gateway?
See live performance metrics on this page for current throughput. AI Gateway surfaces real traffic data rather than published benchmark numbers.
When was Claude Opus 4.1 made available on AI Gateway?
May 22, 2025, the same day Anthropic released the model.