Anthropic launches Claude Opus 5.5 with lower agent costs
Anthropic says Claude Opus 5.5 improves agentic coding and knowledge work while cutting typical Opus 5 workload costs by 40%.
AI-generated: This article was created and published automatically by LinkLoot and was not substantively reviewed by a human editor.
Anthropic has introduced Claude Opus 5.5, the first release in a new Claude 5.5 family. The company positions it as its leading model for agentic coding, computer use, and knowledge work, with a claimed 40% reduction in typical workload cost compared with Claude Opus 5.
The launch matters for teams running long coding or research agents: Anthropic says Opus 5.5 is faster, uses fewer tokens, and raises five-hour usage limits on Pro, Max, Team, and seat-based Enterprise plans. Subscription users also receive a rate-limit reset that can be saved and used when needed.
Claude Opus 5.5 targets long-running engineering work
Anthropic highlights codebase-wide migrations, audits, software optimization, and multi-step computer use. In the company's testing, one external evaluator completed a 680,000-line code migration in less than a day. Another reported that a 200,000-line audit took under three hours with Opus 5.5, compared with more than 20 hours using Opus 5.
Those are vendor-reported evaluations, not independent reproduction. Artificial Analysis independently lists Opus 5.5 as a text-and-image-input model with a 1-million-token context window and a score of 58 on its Intelligence Index. That page also lists the same headline API price of $4 per million input tokens and $20 per million output tokens.
Pricing falls while access expands
Anthropic lists Opus 5.5 at $4 per million input tokens, $20 per million output tokens, $0.20 per million cache reads, and $5 per million cache writes. Compared with Opus 5, the listed input and output prices are 20% lower, while cache reads are 60% lower. Anthropic says the combination of lower token prices and fewer tokens per task produces a 40% typical workload-cost reduction.
Fast mode is available in Claude Code and the Claude Platform at up to 2.5× speed, priced at $8 per million input tokens and $40 per million output tokens. Builders should therefore evaluate both standard and fast-mode economics against their own latency and token-use patterns.
Safety controls remain part of the rollout
Anthropic says Opus 5.5 is less likely than recent models to take hard-to-reverse actions or cross assigned boundaries, and that it is more resistant to prompt injection than Opus 5 in the company's tests. The model is launching with safeguards similar to those used for Anthropic's most capable models.
The company is limiting some high-risk use cases: vetted organizations can apply for a Life Sciences Verification Program, while expanded access to a Cyber Verification Program is planned for the coming weeks. Those programs are not general availability and should not be treated as unrestricted access to biology or cybersecurity workflows.
What builders should check now
- Compare Opus 5.5 with the current production model on real repository migrations, tool-call reliability, and total task cost.
- Recalculate cache-heavy agent workloads; the cache-read price change may matter more than the headline token price.
- Verify plan limits, reset behavior, fast-mode pricing, and API availability in the account that will run the workload.
- Read the Opus 5.5 system-card details before enabling autonomous actions or sensitive research tasks.
Anthropic says Claude Sonnet 5.5 and Claude Haiku 5.5 will follow in the coming weeks. For now, the concrete change is Opus 5.5's combination of frontier positioning, lower stated agent costs, expanded subscription limits, and a staged rollout for sensitive domains.
Sources
Try the related loot
Give Any Model a Sandboxed Shell and File Workspace with OpenRouter