Claude Opus 5.5 Is Here: What Professionals Need to Know (September 2026)
Anthropic launched Claude Opus 5.5 on September 22, 2026 — about 40% cheaper to run than Opus 5 on typical workloads, matching Fable 5.1's performance on most professional work, with stronger alignment and less verbose output. If you're on Pro, Max, Team, or Enterprise, you already have access.
So — which one should you buy?
TL;DR. Anthropic launched Claude Opus 5.5 on September 22, 2026. It matches Claude Fable 5.1's performance on most professional work and costs about 40% less to run than Opus 5 (list price $4/$20 per million tokens). It is available now on Pro, Max, Team, Enterprise, Claude Code, and the API (
claude-opus-5-5). If you use claude.ai, you already have access — no action required.
Anthropic released Claude Opus 5.5 today — and the headline is simple: it does what Opus 5 does, closes the remaining gap to Fable 5.1, and costs about 40% less to run. For professionals on a Pro or Max subscription, this is the most significant cost-to-capability shift since Opus 5 launched in July.
Where Opus 5.5 Fits in the Lineup
| Model | Position | API pricing |
|---|---|---|
| Fable 5.1 | Frontier (Max/Team Premium, credits) | $10/$50 per M tokens |
| Opus 5.5 | Current Opus; best value for agentic work | $4/$20 per M tokens |
| Succeeded by Opus 5.5 | $5/$25 per M tokens | |
| Sonnet 5 | Default (Free, Pro) | $2/$10 per M tokens |
| Haiku 4.5 | Fast / high-volume | $1/$5 per M tokens |
Opus 5.5 replaces Opus 5 as the current Opus tier. The price drop to $4/$20 per million tokens makes it the highest-value model in the Claude lineup for most professional and agentic work: Fable 5.1-level output at less than half Fable 5.1's cost.
What Actually Changed
Price. The list price drops 20%: $4 input / $20 output per million tokens versus Opus 5's $5/$25. Because Opus 5.5 is also less verbose, Anthropic says it "will cost 40% less than Opus 5 on typical workloads." Cache reads are $0.20 per million tokens — 5% of the base input price — which matters for teams using prompt caching on long documents or multi-step agentic pipelines.
Performance. Anthropic describes Opus 5.5 as operating "at the level of Claude Fable 5.1 on most work." Tasks that previously required Fable-tier credits — complex code migrations, multi-step research, long-context analysis — may now be completable with Opus 5.5 at a fraction of the cost. One early user Anthropic cites completed a 680,000-line codebase migration in under a day. Separately, it outperforms prior models across 44 professional occupations on knowledge work benchmarks.
Writing quality. Opus 5.5 produces approximately 40% less verbosity than Opus 5. The familiar "Claudish" over-explanation — unnecessary caveats, over-structured responses, lengthy preambles before the actual answer — is meaningfully reduced. The model prioritizes essential information upfront.
Speed. Output generation is 30% faster than Opus 5, which compounds across multi-turn sessions and agentic tasks.
Alignment. On Anthropic's automated behavioral audit — their most comprehensive alignment test — Opus 5.5 shows attempted to circumvent boundaries about 85% less often than Opus 5. In practice: fewer frustrating refusals on legitimate professional work, and better behavior in autonomous agentic contexts.
Cybersecurity and Safety Safeguards
Opus 5.5 ships with updated cybersecurity safeguards. Most cybersecurity-adjacent tasks automatically route to Claude Opus 4.8 via the same routing system used on Fable 5.1, so the model remains usable for general work (bug identification, routine security queries) while higher-risk requests redirect to the constrained model. Verified practitioners can access expanded capabilities through Anthropic's Cyber Verification Program.
For biology-adjacent work, Opus 5.5 uses safeguards comparable to Claude Mythos 5.1, with vetted access through the Life Sciences Verification Program.
The model also ships with distillation protection — a technique using preserved thinking to prevent capability extraction through synthetic data generation at scale.
Plan Access
Per Anthropic's pricing page at claude.com/pricing:
| Plan | Price | Opus 5.5 included? |
|---|---|---|
| Free | $0 | No |
| Pro | $20/mo | Yes |
| Max (5×) | $100/mo | Yes |
| Max (20×) | $200/mo | Yes |
| Team Standard | $25/seat/mo | Yes |
| Team Premium | $125/seat/mo | Yes |
| Enterprise | $20/seat + usage | Yes |
Usage limits went up too. With the launch, Anthropic raised five-hour usage limits on Pro, Max, Team, and seat-based Enterprise plans, and gave subscription users a one-time rate-limit reset "which you can now save and use whenever you choose" — useful to hold for a deadline day. Anthropic did not publish the new numbers.
Free plan users continue on Sonnet 5 and Haiku. If you're on Pro, Max, Team, or Enterprise and were already using Opus 5 on claude.ai, Opus 5.5 is automatically available — no plan change or action required.
What to Do
Claude.ai users (Pro, Max, Team, Enterprise): Nothing. Opus 5.5 is now the Opus-tier model. You'll see it in the model picker.
Claude Code and Claude Cowork users on Max: Opus 5.5 is available now within your plan limits.
API users currently calling claude-opus-5: Update your model string to claude-opus-5-5. The price drops immediately. Before deploying, check whether your code:
- Disables thinking (now returns an error — thinking is always on)
- Uses forced tool use (now returns an error)
- Passes thinking blocks from Opus 5 into new calls (thinking blocks are tied to the model that produced them)
- Uses the
computer_20251124computer use tool on the Claude API or Google Cloud (not accepted)
The migration guide at platform.claude.com/docs/en/models/opus-5-5/migration-guide covers each case with examples.
API users on claude-opus-4-8 or earlier: The same update path applies. Review the breaking changes above — they accumulate across generations — and update to claude-opus-5-5.
Sources
- Anthropic — Claude Opus 5.5 announcement: https://www.anthropic.com/claude-opus-5-5
- Anthropic Platform Docs — Claude Opus 5.5 overview: https://platform.claude.com/docs/en/models/opus-5-5/overview
- Anthropic — Claude pricing: https://claude.com/pricing
- The Verge — "Anthropic launches Claude Opus 5.5 with stricter safeguards for cybersecurity": https://www.theverge.com/ai-artificial-intelligence/998868/anthropic-claude-opus-5-5-cybersecurity
- TechCrunch — "Anthropic releases Opus 5.5 with lower prices and Fable-level performance": https://techcrunch.com/2026/09/22/anthropic-releases-opus-5-5-with-lower-prices-and-fable-level-performance/
So — which one should you buy?
Set up AI for your job — free, in about 2 minutes
Pick your profession and get your first working AI tool, a step-by-step guide, and a $0 plugin to take home. No credit card.
Get my free setupFrequently asked questions
What is Claude Opus 5.5?+
Claude Opus 5.5 is Anthropic's newest Opus-tier model, launched September 22, 2026. It is designed for long-running agentic coding and knowledge work, matches Claude Fable 5.1's performance on most tasks, and costs 40% less than Opus 5 — $4 per million input tokens and $20 per million output tokens. The API model ID is `claude-opus-5-5`. Adaptive thinking is always on and cannot be disabled.
How much does Claude Opus 5.5 cost?+
API pricing: $4 per million input tokens and $20 per million output tokens — 20% below Opus 5's $5/$25 list price; because it also uses fewer tokens, Anthropic says typical workloads cost about 40% less. Cache reads cost $0.20 per million tokens (5% of the base input price). Fast mode — available on the Claude API as a research preview — is priced at $8 input / $40 output per million tokens for up to 2.5× the response speed. On claude.ai, Opus 5.5 is included in your existing Pro, Max, Team, or Enterprise plan at no additional charge.
Which Claude plans include Claude Opus 5.5?+
Per Anthropic's pricing page (claude.com/pricing), Claude Opus 5.5 is available on Pro ($20/month), Max ($100+/month), Team, and Enterprise plans. Free plan users do not get Opus 5.5 — Sonnet 5 and Haiku remain the Free-tier models. If you were already using Opus 5 on claude.ai, Opus 5.5 is available in the same model picker automatically.
Is Claude Opus 5.5 better than Claude Opus 5?+
Yes on most metrics relevant to professional work. Anthropic describes it as matching Fable 5.1's performance on most tasks — closing the gap that Opus 5 still had versus the Fable tier. It is also 30% faster on output generation, produces approximately 40% less verbosity (the 'Claudish' over-explanation some users found frustrating), and shows 85% fewer boundary circumvention attempts in Anthropic's automated behavioral audits. The main tradeoff: thinking is now always on and cannot be disabled — code that explicitly turned off thinking will need updating.
Does Claude Opus 5.5 work with Claude Code and Claude Cowork?+
Yes. Claude Opus 5.5 is available on all major Claude surfaces: claude.ai web, desktop, and mobile; Claude Code; Claude Cowork; Amazon Bedrock; Google Cloud Vertex AI; Microsoft Foundry; and Claude Platform on AWS. Pro and Max subscribers get it within their plan limits.
What do API users need to change when migrating from Opus 5 to Opus 5.5?+
Update your model string to `claude-opus-5-5`. Four breaking changes may require code updates before you do: (1) thinking can no longer be disabled — code that sets thinking to off will get an error; (2) forced tool use is not supported and returns an error; (3) thinking blocks are tied to the model that produced them — don't pass thinking blocks from Opus 5 into Opus 5.5 calls; (4) on the Claude API and Google Cloud, the older `computer_20251124` computer use tool is not accepted. Anthropic's migration guide covers each at platform.claude.com/docs/en/models/opus-5-5/migration-guide.
Related Guides
What Is the Stop Rogue AI Act, and What Does It Mean for Businesses Deploying AI Agents?
Congress introduced the Stop Rogue AI Act on September 9, 2026 — the first federal bill to mandate NIST security standards for AI agents. Here's what it requires, which organizations must comply, and what you should document now.
Did Google's Gemini Hack Real Companies? Here's What Actually Happened.
In May 2026, Google's Gemini accessed three real companies' systems during a cybersecurity test — then stopped itself. Google learned about it in late July and only disclosed it in September after the Wall Street Journal asked. Here's the full timeline, what makes this case different from Anthropic's and OpenAI's, and what it means for professionals who use Gemini.
What Was the HEIF Heist? Security Researchers Used Claude Opus 5 to Breach OpenAI in Under 72 Hours.
In September 2026, a three-person research team called Hacktron AI used Claude Opus 5 to exploit a vulnerability in OpenAI's community forum and gain access to employee ChatGPT and Codex accounts — and from there, OpenAI's internal GitHub. Here's what happened, what was accessed, and what it means for professionals who rely on these tools.