# Claude Sonnet 4.5

Claude Sonnet 4.5 advanced multiple capability areas simultaneously: strong results on real-world software engineering benchmarks, strong computer-use benchmark results at release, and the ability to sustain over 30 hours of autonomous agentic coding.

- **Model ID:** `anthropic/claude-sonnet-4.5`
- **Type:** chat
- **Providers:** anthropic, bedrock, vertexAnthropic
- **Context window:** 1,000,000
- **Maximum output tokens:** 64,000
- **Pricing:** $3.0/1M input tokens, $15.0/1M output tokens
- **Canonical page:** https://vercel.com/ai-gateway/models/claude-sonnet-4.5

## Supported parameters

Detailed capability metadata has not been reported for this model.

## Example

```ts
import { streamText } from 'ai'

const result = streamText({
  model: 'anthropic/claude-sonnet-4.5',
  prompt: 'Why is the sky blue?'
})
```

## About

Claude Sonnet 4.5 launched on September 29, 2025. The OSWorld result backed Anthropic's computer use claims directly: 61.4%, up from Sonnet 4's 42.2% four months earlier. On SWE-bench Verified, Claude Sonnet 4.5 scored 77.2% and maintained focus for 30+ hours on complex multi-step tasks, a duration threshold that changes what's architecturally feasible for autonomous engineering work.

Domain expert evaluation reinforced the benchmark numbers. Finance, law, medicine, and STEM specialists found substantially better domain-specific knowledge and reasoning compared to older models including Opus 4.1. Devin increased planning performance by 18% and end-to-end scores by 12%, the biggest jump since Claude Sonnet 3.6. Cursor, GitHub Copilot, and Figma Make reported significant gains in their specific domains. Claude Code shipped checkpoints and rollback, a native VS Code extension, and a refreshed terminal interface alongside this model.

At release, Claude Sonnet 4.5 included substantial alignment improvements over prior Claude models. Safety gains are concrete: substantial reductions in sycophancy, deception, power-seeking, and tendency to encourage delusional thinking. Prompt injection defense for computer use and agentic capabilities improved considerably. Anthropic released Claude Sonnet 4.5 under ASL-3 (AI Safety Level 3) protections, the first Claude model at that safety level, with CBRN (chemical, biological, radiological, and nuclear) classifiers active.

The Claude Agent SDK launched alongside Claude Sonnet 4.5, giving you access to the same infrastructure that powers Claude Code: memory management, permission systems, and subagent coordination for building custom agents.

## What to consider

Claude Sonnet 4.5's computer use capability is protected by ASL-3 (AI Safety Level 3) safeguards: classifiers that screen for potentially dangerous inputs and outputs. These may occasionally flag normal content. Anthropic has substantially reduced false positive rates since the classifiers were first deployed.

## When to use

### Best For

- **Computer use and real-world browser/software automation**, with strong results on OSWorld at release among models evaluated then
- **Extended autonomous coding sessions**, with documented 30+ hour capability for complex multi-step engineering tasks
- **Complex agent workflows** where Anthropic explicitly positioned it for agent workloads at release
- **Domain-specific applications in finance, law, medicine, and STEM** where expert evaluation showed substantial gains
- **Production deployments requiring strong alignment properties**, with reduced sycophancy and deception compared to earlier Claude releases at that time

### Consider Alternatives When

- Cost is the primary constraint. Check Haiku 4.5 for workloads where its capability-per-cost profile is sufficient
- The task is simple and latency-sensitive. Sonnet 4.5's capability depth comes with higher per-token cost than lighter models
- You need the context window of 1M tokens at Sonnet tier. Check if Claude Sonnet 4.6 covers both requirements
- The specific computer use or coding task falls within what earlier models handle equivalently

## Best for

- **Computer use and software automation:** Strong results on OSWorld at release among models evaluated then
- **Extended autonomous coding sessions:** Documented 30+ hour capability for complex multi-step engineering tasks
- **Complex agent workflows:** Anthropic explicitly positioned Claude Sonnet 4.5 for agent workloads at release
- **Finance, law, medicine, and STEM applications:** Expert evaluation showed substantial gains in domain knowledge and reasoning compared to Opus 4.1
- **Strong alignment properties:** Reduced sycophancy and deception compared to earlier Claude releases at that time

## Consider alternatives

- **Primary cost constraint:** Haiku 4.5 may offer sufficient capability-per-cost for lighter workloads
- **Simple latency-sensitive tasks:** Claude Sonnet 4.5's capability depth comes with higher per-token cost than lighter models
- **Sonnet-tier large context:** Claude Sonnet 4.6 covers both 1M tokens context and Sonnet pricing
- **Earlier-model parity:** Earlier models handle some specific computer use or coding tasks equivalently

## Frequently asked questions

### What was Claude Sonnet 4.5's OSWorld score and why does it matter?

61.4%, up from Sonnet 4's 42.2% four months earlier. OSWorld measures AI performance on real-world computer tasks: navigating software, filling forms, and clicking UI elements. It focuses on operational computer-use scenarios rather than abstract reasoning alone.

### How long can Claude Sonnet 4.5 maintain focus on a single agentic coding task?

More than 30 hours on complex, multi-step tasks. Anthropic noted this duration changes what's architecturally feasible for autonomous engineering work. Individual results vary by task structure.

### What is ASL-3 and why does it apply to Sonnet 4.5?

ASL-3 (AI Safety Level 3) is Anthropic's framework level for models requiring additional safeguards. Sonnet 4.5 is the first Claude model released under ASL-3 protections, which include classifiers screening inputs and outputs for CBRN-related content. False positive rates have decreased by a factor of 10 since initial deployment.

### What is the Claude Agent SDK and how does it relate to this model?

The Claude Agent SDK launched alongside Sonnet 4.5. It gives you access to the same agent infrastructure that powers Claude Code: memory management across long tasks, permission systems, and subagent coordination. Use it to build custom agents on the same foundation.

### What alignment improvements came with Sonnet 4.5?

Substantial reductions in sycophancy, deception, power-seeking, encouragement of delusional thinking, and compliance with harmful system prompts, measured via an automated behavioral auditor. The model also improved defenses against prompt injection attacks for computer use and agentic capabilities.

### Why did specialists in finance, law, medicine, and STEM find Sonnet 4.5 significantly better than previous models?

Professionals assessed domain-specific knowledge and reasoning in Anthropic's expert evaluations. Results showed substantially better performance compared to older models, including Opus 4.1. The intelligence improvements extend beyond coding benchmarks.

### Is Sonnet 4.5 priced differently from Sonnet 4?

Current pricing is shown on this page. AI Gateway routes across providers, and rates may vary by provider.

## Links

- [Model page](https://vercel.com/ai-gateway/models/claude-sonnet-4.5)
- [AI Gateway documentation](https://vercel.com/docs/ai-gateway)
- [Provider model documentation](https://www.anthropic.com/claude/sonnet)
- [Provider pricing](https://www.anthropic.com/pricing#anthropic-api)


If you encounter incorrect information, can’t find what you need, need a capability Vercel doesn’t currently offer, or have any other feedback to report, please do it by following the instructions at https://vercel.com/feedback.md.