Model comparison
Claude Opus 5.5 vs Claude Opus 5: is it worth switching?
Claude Opus 5.5 is 20% cheaper per token than Claude Opus 5, and cache hits cost 60% less. What that means for Claude Code sessions, and what stays the same.
· Prices as of September 26, 2026
Claude Opus 5.5
Anthropic · Released September 22, 2026
Anthropic's recommended starting model for most work, built for long-running agentic coding. Anthropic says it matches Claude Fable 5.1 on most work at a much lower price.
Claude Opus 5.5 facts and comparisonsClaude Opus 5
Anthropic · Released July 24, 2026 · Previous generation
The previous everyday Opus, pitched as close to Claude Fable 5 at half the price. Anthropic now recommends moving to Claude Opus 5.5.
Claude Opus 5 facts and comparisons
The short answer
Claude Opus 5.5 replaces Claude Opus 5 as Anthropic's recommended Opus and as Claude Code's opus default, at 20% lower list prices and a cache hit that costs $0.20 instead of $0.50. On the example agentic coding session that is $4.40 against $6.00, 27% less. Opus 5 stays available as a legacy model, but it costs more on every line of the rate card.
Choose Claude Opus 5.5 if
- You want Claude Code's opus default, which moved from Opus 5 to Opus 5.5.
- Your sessions lean on the cache, where hits are 60% cheaper than on Opus 5.
- You use fast mode, which costs $8 input and $40 output per million on Opus 5.5 against $10 and $50 on Opus 5.
- You adjust effort during a task: Anthropic notes that changing effort on Opus 5.5 keeps the prompt cache.
Choose Claude Opus 5 if
- You have prompts, evaluations, or agents tuned against Opus 5 and need time to re-validate them on Opus 5.5.
- You call Opus 5 by its model id in a pipeline and want to switch on your own schedule, since Anthropic keeps it available.
Side by side
Specs and prices
| Fact | Claude Opus 5.5 | Claude Opus 5 |
|---|---|---|
| Maker | Anthropic | Anthropic |
| API model id | claude-opus-5-5 | claude-opus-5 |
| Released | September 22, 2026 | July 24, 2026 |
| Status | Current | Previous generation |
| Context window | 1M tokens | 1M tokens |
| Max output | 128K tokens | 128K tokens |
| Open weights | No | No |
| Input, per 1M tokens | $4 | $5 |
| Cache hit, per 1M | $0.20 | $0.50 |
| Cache write, per 1M | $5 (5-minute), $8 (1-hour) | $6.25 (5-minute), $10 (1-hour) |
| Output, per 1M tokens | $20 | $25 |
| Runs in | Claude Code, Cursor, OpenCode, OpenRouter, and GitHub Copilot | Claude Code, Cursor, OpenCode, OpenRouter, and GitHub Copilot |
Standard API rates in US dollars, as published by each maker on September 26, 2026. Batch and priority tiers, taxes, and subscription plans are not included. Claude Opus 5.5: The full 1M context window is billed at standard rates. Fast mode, a research preview on the Claude API, costs $8 input and $40 output per million tokens. Claude Opus 5: The full 1M context window is billed at standard rates. Fast mode, a research preview on the Claude API, costs $10 input and $50 output per million tokens.
Cost
What the same work costs
The same token counts, priced at each model’s published rates. Cached tokens are billed at each maker’s cache prices, so the session shows what caching is worth on each model.
Real sessions differ: the two models count the same code as different numbers of tokens, and reasoning settings change how much each one writes. Your own history is the real test.
| Workload | Claude Opus 5.5 | Claude Opus 5 |
|---|---|---|
| Agentic coding session, 100K input, 400K written to cache, 2M read from cache, 50K output | $4.40 | $6.00 |
| Large one-off review, 150K input with no cache hits, 10K output | $0.80 | $1.00 |
| Output-heavy generation, 30K input, 80K output | $1.72 | $2.15 |
| A month of sessions, 110 sessions: 5 a day, 22 working days | $484.00 | $660.00 |
| Where the session’s cost goes | ||
| Cache writes | $2.60 | $3.25 |
| Cache reads | $0.40 | $1.00 |
| Uncached input | $0.40 | $0.50 |
| Output | $1.00 | $1.25 |
- caching saves on the session with Claude Opus 5.5 (60%)
- $6.60
- caching saves on the session with Claude Opus 5 (56%)
- $7.75
How much does Claude Opus 5.5 save over Claude Opus 5?
Claude Opus 5.5 lists at $4 per million input tokens and $20 per million output tokens. Claude Opus 5 lists at $5 and $25. That is 20% less on input, output, and both kinds of cache write: $5 and $8 on Opus 5.5 for 5-minute and 1-hour writes, against $6.25 and $10. Uncached work shows that 20% directly, with the large one-off review at $0.80 against $1.00.
The cache hit changed more than the rest. Opus 5 bills a hit at 0.1x input, $0.50 per million, while Opus 5.5 bills at 0.05x, $0.20 per million, which Anthropic describes as cache reads 60% cheaper. In the example session the 2M cached tokens cost $0.40 instead of $1.00. That $0.60 is why the session gap, 27%, is wider than the 20% list gap: $4.40 against $6.00.
Over 110 sessions a month the difference is $176.00, $484.00 against $660.00 at API rates. The largest line on both is still cache writes, $2.60 on Opus 5.5 and $3.25 on Opus 5, because both bill a 1-hour write at 2x input.
Anthropic says 40% cheaper to run. Why does this page show 27%?
Anthropic's launch claim is, "It performs at the level of Claude Fable 5.1 on most work and costs 40% less to run than Opus 5." The figures here hold the token counts fixed and change only the rates, which gives 27% less on the session and 20% less on uncached work.
A model's running cost depends on how many tokens it thinks and writes, not only on its rates, and the rates alone can't reproduce Anthropic's 40%. The dependable way to see the difference on your own work is to compare the same kind of task on both models in your own history. EveryToken gives you the numbers from your local Claude Code log: it prices each request at API rates and splits cost and cache savings by model.
What else changes when you move to Opus 5.5
Very little in the spec sheet. Both have a 1M context window billed at standard rates across the full window, both write up to 128K tokens of output, and both use Anthropic's newer tokenizer, which counts about 30% more tokens than earlier Claude models for the same text. Moving between them doesn't change how many tokens a prompt counts as.
Anthropic also says Opus 5.5 outputs more than 30% faster than Opus 5, and it names long, sprawling jobs such as codebase-wide migrations and audits as a particular strength. Opus 5 was Claude Code's opus default until Opus 5.5 replaced it, and Anthropic now recommends the move. Both are listed in Claude Code, Cursor, OpenRouter, OpenCode, and GitHub Copilot.
Opus 5.5 has adaptive thinking always on, with medium as the default effort, and changing its effort keeps the prompt cache. That means you can raise effort for a hard step in a long session without paying to write the context to the cache again.
Prompt caching
How Anthropic bills cached tokens
Anthropic
Claude caches a prompt prefix up to a breakpoint. One top-level cache_control field places the breakpoint automatically and moves it as the conversation grows, or you can mark up to 4 blocks yourself. Claude Code manages caching for you.
A 5-minute cache write costs 1.25x the input price and a 1-hour write costs 2x. A cache hit costs 0.1x input on most models, 0.05x on Claude Opus 5.5, and 0.025x on Claude Fable 5.1 and Claude Mythos 5.1. Every hit restarts the cache lifetime at no charge.
Anthropic's rule of thumb: a 5-minute write pays for itself after one cache read, and a 1-hour write after two.
In Claude Code, the main conversation uses the 1-hour cache on a Claude subscription and the 5-minute cache with an API key. Each model has its own cache, so switching models starts over.
Source: Anthropic: Prompt caching
Your own numbers
See what Claude Opus 5.5 and Claude Opus 5 really cost you.
everyaitoken reads your Claude Code, Cursor, OpenCode, and OpenRouter history on your Mac and prices every request at API rates, with what caching saved or cost. $9 once.
FAQ
Questions
Is Claude Opus 5.5 cheaper than Claude Opus 5?
Yes, on every line of the rate card. Input, output, and cache writes are 20% cheaper, and cache hits are 60% cheaper. The example agentic session costs $4.40 on Opus 5.5 and $6.00 on Opus 5.
Is Claude Opus 5 still available?
Yes. Anthropic keeps it as a legacy model and recommends moving to Opus 5.5. It was Claude Code's opus default before Opus 5.5 took over.
Will my prompts use more tokens on Claude Opus 5.5?
Not because of the tokenizer, since both models use Anthropic's newer one. Output length can still differ, because effort settings and the model itself shape how much each response writes.
Does Claude Opus 5.5 keep the 1M context window?
Yes. Both models accept 1M tokens of context at standard rates, with no long-context surcharge, and both write up to 128K tokens of output.
Sources
- Anthropic: Pricing
- Anthropic docs: Claude Opus 5.5
- Anthropic: Claude Opus 5.5
- Anthropic docs: Fast mode
- Claude Code docs: Model configuration
- Cursor docs: Claude Opus 5.5
- OpenRouter: Claude Opus 5.5
- OpenCode docs: Zen
- GitHub Docs: Supported AI models in Copilot
- Anthropic docs: Claude Opus 5
- Anthropic: Introducing Claude Opus 5
- Cursor docs: Claude Opus 5
- OpenRouter: Claude Opus 5
- Anthropic: Prompt caching