Model comparison
Claude Opus 5.5 vs Claude Sonnet 5: cost and caching
Claude Opus 5.5 costs twice as much per token as Claude Sonnet 5, yet cache hits cost $0.20 on both. What that means for agentic coding sessions.
· 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 Sonnet 5
Anthropic · Released June 30, 2026
Anthropic's balance of speed and intelligence, and a drop-in upgrade for Claude Sonnet 4.6. Anthropic pitches it as close to Claude Opus 4.8 at a lower price.
Claude Sonnet 5 facts and comparisons
The short answer
Claude Sonnet 5 costs half as much per token as Claude Opus 5.5, but cache hits cost $0.20 per million on both. On a cache-heavy agentic coding session that narrows the gap to 1.8x, $2.40 against $4.40. Opus 5.5 is Anthropic's pick for long, sprawling jobs, and Sonnet 5 is the cheaper everyday choice with the same 1M context window.
Choose Claude Opus 5.5 if
- You run long agentic jobs such as codebase-wide migrations and audits, which Anthropic names as a particular strength of Opus 5.5.
- You use Claude Code on a Pro, Max, Team, or Enterprise plan and want its default model.
- You want fast mode, a research preview on the Claude API at $8 input and $40 output per million tokens.
- Your sessions lean on the cache: a hit costs 5% of the Opus 5.5 input price, so long sessions cost less than the headline 2x gap suggests.
Choose Claude Sonnet 5 if
- You want the lower price for everyday coding: $2 input and $10 output per million tokens, half the Opus 5.5 rates.
- You send many short, uncached requests, where the full 2x price gap applies.
- You need the same 1M context window and 128K of output at a lower cost.
- You are moving up from Claude Sonnet 4.6, since Anthropic calls Sonnet 5 a drop-in upgrade.
Side by side
Specs and prices
| Fact | Claude Opus 5.5 | Claude Sonnet 5 |
|---|---|---|
| Maker | Anthropic | Anthropic |
| API model id | claude-opus-5-5 | claude-sonnet-5 |
| Released | September 22, 2026 | June 30, 2026 |
| Status | Current | Current |
| Context window | 1M tokens | 1M tokens |
| Max output | 128K tokens | 128K tokens |
| Open weights | No | No |
| Input, per 1M tokens | $4 | $2 |
| Cache hit, per 1M | $0.20 | $0.20 |
| Cache write, per 1M | $5 (5-minute), $8 (1-hour) | $2.50 (5-minute), $4 (1-hour) |
| Output, per 1M tokens | $20 | $10 |
| 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 Sonnet 5: The full 1M context window is billed at standard rates. The launch price became the standard price on August 10, 2026, and a planned increase was cancelled.
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 Sonnet 5 |
|---|---|---|
| Agentic coding session, 100K input, 400K written to cache, 2M read from cache, 50K output | $4.40 | $2.40 |
| Large one-off review, 150K input with no cache hits, 10K output | $0.80 | $0.40 |
| Output-heavy generation, 30K input, 80K output | $1.72 | $0.86 |
| A month of sessions, 110 sessions: 5 a day, 22 working days | $484.00 | $264.00 |
| Where the session’s cost goes | ||
| Cache writes | $2.60 | $1.30 |
| Cache reads | $0.40 | $0.40 |
| Uncached input | $0.40 | $0.20 |
| Output | $1.00 | $0.50 |
- caching saves on the session with Claude Opus 5.5 (60%)
- $6.60
- caching saves on the session with Claude Sonnet 5 (56%)
- $3.10
Why the price gap shrinks on agentic sessions
On paper the gap is simple. Claude Opus 5.5 charges $4 per million input tokens and $20 per million output tokens, exactly twice the $2 and $10 of Claude Sonnet 5. Cache writes follow the same ratio: $5 and $8 per million for 5-minute and 1-hour writes on Opus 5.5, against $2.50 and $4 on Sonnet 5.
Cache reads break the pattern. Both models charge $0.20 per million cached tokens, because Anthropic prices a cache hit on Opus 5.5 at 5% of its input price instead of the usual 10%. In the example session, where 2M input tokens come from the cache, reads cost $0.40 on either model.
The result is a 1.8x gap on the session, $4.40 against $2.40, compared with 2x on the uncached review and on the output-heavy generation. The more of its context a coding tool resends from the cache, the closer the two models get in cost.
Cache writes are the biggest line item
In the session, cache writes cost $2.60 on Opus 5.5 and $1.30 on Sonnet 5, which is 59% and 54% of each total. Most of that comes from the 1-hour writes, which Anthropic bills at 2x the input price.
This matters in Claude Code. The main conversation uses the 1-hour cache on a Claude subscription and the 5-minute cache with an API key. Anthropic's rule of thumb is that a 5-minute write pays for itself after one cache read and a 1-hour write after two, and that holds on both models.
Caching saves more dollars on Opus 5.5 because its input is dearer: $6.60 on the session, or 60% of what the same tokens would cost uncached, against $3.10 and 56% on Sonnet 5. The cache math on our homepage walks through the same session line by line.
What Anthropic says each model is for
Anthropic calls Opus 5.5 its recommended starting model for most work and says it performs at the level of Claude Fable 5.1 on most work. It names long, sprawling jobs like codebase-wide migrations and audits as a particular strength, and Opus 5.5 is the default model in Claude Code on Pro, Max, Team, and Enterprise plans.
Sonnet 5 is Anthropic's balance of speed and intelligence. Anthropic describes its performance as close to Claude Opus 4.8 at lower prices. Both models share a 1M context window with no long-context premium, 128K of output, and Anthropic's newer tokenizer, so the same prompt counts as the same number of tokens on either.
Their defaults differ. Opus 5.5 runs at medium effort by default and Sonnet 5 at high, and effort changes how many tokens each one writes. Compare them on your own tasks before moving a whole team, and read the result from your own history rather than a price sheet.
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 Sonnet 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 twice as expensive as Claude Sonnet 5?
Per token, yes: $4 against $2 for input and $20 against $10 for output. Cache hits cost $0.20 on both, so a cache-heavy agentic session costs 1.8x as much on Opus 5.5 rather than 2x. At 110 sessions a month, the example session comes to $484.00 on Opus 5.5 and $264.00 on Sonnet 5.
Which model does Claude Code use by default?
Claude Opus 5.5 is the default in Claude Code on Pro, Max, Team, and Enterprise plans and with an Anthropic API key. The sonnet alias resolves to Claude Sonnet 5 on the Anthropic API, so you can switch per session.
Do both models have a 1M token context window?
Yes. Both accept 1M tokens of context at standard rates, with no long-context premium, and both write up to 128K tokens of output.
How can I see what each model costs me?
Claude Code records the tokens every request used in its local history. EveryToken reads that history on your Mac and prices each request at Anthropic's rates, split by model, with what caching saved or cost.
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 Sonnet 5
- Anthropic: Introducing Claude Sonnet 5
- Cursor docs: Claude Sonnet 5
- OpenRouter: Claude Sonnet 5
- Anthropic: Prompt caching