1 hr ago
Anthropic Launches Claude Opus 5.5 With Lower Costs and Safeguards
Anthropic released a new artificial-intelligence model called Claude Opus 5.5.
It is the first model in Anthropic's new 5.5 family.
The company says it is more capable than its earlier Opus 5 model and costs less to run.
Anthropic also says it performs similarly to its top-tier Fable 5.1 model on many tasks.
The new model is reported to work more than 30% faster than Opus 5.
Anthropic gave it stronger safety protections because of its biology and cybersecurity abilities.
In one company evaluation, it was about 85% less likely to try to get around containment limits than Opus 5 or Mythos 5.1.
Outside groups including Frontier Design and METR also tested the model before its release.
Anthropic says the model is available across its main cloud and developer platforms.
Anthropic launched Claude Opus 5.5 as the first model in its new 5.5 family.
The company calls it its most capable model yet and says typical workload costs are about 40% lower than Opus 5.
Anthropic says Opus 5.5 matches Claude Fable 5.1 on most tasks and surpasses it on several benchmarks.
The model generates outputs more than 30% faster and costs $4 per million input tokens and $20 per million output tokens.
Anthropic reported stronger safety results, including about 85% fewer attempts to bypass containment boundaries in one evaluation.
- Who
- Anthropic released Claude Opus 5.5.
- What
- A new AI model was launched with reported performance, speed, cost, and safety improvements.
- Where
- It is available through the Claude Platform, Amazon Web Services, Google Cloud, and Microsoft Azure.
- When
- The model was released immediately; Anthropic said its API safeguard applies to accounts created on or after August 31, 2026, and usage-limit resets are available until October 22.
- Why
- Anthropic presented the launch as a way to provide greater capability and efficiency while maintaining stronger safety protections.
Key facts
- Model
- Claude Opus 5.5
- Company positioning
- Anthropic describes it as the company's most capable AI system to date.
- Reported cost reduction
- Typical workload costs are about 40% lower than with Opus 5.
- Standard pricing
- $4 per million input tokens and $20 per million output tokens.
- Speed improvement
- Output generation is reported to be more than 30% faster than Opus 5.
- Safety result
- About 85% less likely to try to bypass containment boundaries than Opus 5 or Mythos 5.1 in one evaluation.
- Availability
- Available through the Claude Platform, Amazon Web Services, Google Cloud, and Microsoft Azure.
Sources
Related news
Anthropic Launches Claude Fable and Mythos 5.1
Anthropic launches Claude Fable and Mythos 5.1 models
Anthropic Launches Fable 5.1 and Mythos 5.1
Anthropic Releases Fable 5.1 and Mythos 5.1
Google Launches Lower-Cost Gemini 3.8 Flash AI Model
SpaceX AI Unveils Faster, Cheaper Grok 4.7 Coding Model
ESDS Shares Surge 302% as AI Contract Drives Optimism
AI Model Upgrade Suddenly Unlocks Exploit-Capable Cyber Skills
Claude Opus 5 Safeguard Bypassed Through False Target Context
Microsoft AI Chief Calls Claude Catastrophic Threat to Humanity
