claude-sonnet-5.5 API — per-token pricing and code examples
Price per 1M tokens
Cache rates apply when the provider reports cache tokens.
71% offAbout this model
claude-sonnet-5.5 is served through OpenAI Chat Completions, OpenAI Responses, Anthropic Messages. You pay $0.58 per 1M input tokens and $2.90 per 1M output tokens. That is 71% below the official price of $2.00 and $10.00. Cache reads cost $0.058 per 1M tokens when the provider reports cached tokens. It is included in budget packs from $99.00 up.
Where Claude Sonnet 5.5 fits
Claude Sonnet 5.5 is Anthropic's newest Sonnet, released on September 28, 2026. Anthropic calls it "the best combination of speed and intelligence" and rates its latency as fast. Retirement is listed as not sooner than September 28, 2027, and its knowledge reaches June 2026. It accepts text and images, returns text, and supports 1M tokens of context with up to 128K output tokens in one synchronous reply.
Reach for Sonnet 5.5 when response time matters as much as quality: chat products, coding assistants inside an editor, agents that make many short tool calls, and pipelines that process a large number of requests.
The between_tools setting
Sonnet 5.5 thinks by default, and thinking: {type: "disabled"} returns a 400. What it offers instead is a lighter mode. Anthropic describes between_tools as the lowest thinking setting on this model: as the name says, it turns off the up-front thinking that would otherwise come before the answer. It only works at effort high or below; with xhigh or max the request is rejected. Two related limits apply. The display option is not valid together with between_tools, and once a conversation runs in that mode, changing effort partway through returns a 400.
Effort defaults to high and all five levels are available. Anthropic notes that Sonnet 5.5's levels are recalibrated, so a given level does not produce the same amount of thinking as the same level on Claude Sonnet 5. Re-test your effort choice instead of copying it from an older Sonnet. The budget_tokens form is rejected too.
For tools, forced tool use returns a 400 on Sonnet 5.5; use tool_choice: {"type": "auto"} with strict tool use or structured outputs. The older computer_20251124 computer use tool is not accepted on the Claude API. With the default omitted display, text between tool calls shows up in thinking blocks rather than plain text blocks.
What it costs on Cent AI
Sonnet 5.5 charges your prepaid USD balance per token at the rates shown in the price box. Reasoning counts as output. For tool-heavy agents, between_tools at low or medium effort keeps reasoning output on Sonnet 5.5 as small as it gets, since it is the lowest thinking setting Anthropic offers on this model.
The minimum cacheable prompt on Sonnet 5.5 is 512 tokens. Anthropic's cache lasts five minutes by default, or one hour when you set ttl: "1h", and on Cent AI cache writes for both lifetimes are billed at the input rate. Cache charges follow only the cached token counts that Sonnet 5.5 reports.
Using it from your tools
Sonnet 5.5 is served on /v1/messages, /v1/chat/completions and /v1/responses. Cent AI does not change your parameters silently, so the model's validation decides. Anthropic's Sonnet 5.5 page is blunt: setting temperature, top_p or top_k returns a 400 error. Prefilling the final assistant turn is rejected as well.
A /v1/messages request for a tool-calling agent with up-front thinking off:
{
"model": "claude-sonnet-5.5",
"max_tokens": 2048,
"thinking": { "type": "between_tools" },
"output_config": { "effort": "medium" },
"tools": [{
"name": "search_docs",
"description": "Search the product docs",
"input_schema": {
"type": "object",
"properties": { "query": { "type": "string" } },
"required": ["query"]
}
}],
"messages": [{ "role": "user", "content": "How do I rotate an API key?" }]
}
Anthropic's own API ID is claude-sonnet-5-5 with a hyphen; on Cent AI you send the dotted claude-sonnet-5.5.
Specification
- Context window
- —
- Input and output
- Text
- Endpoints
- /v1/chat/completions, /v1/responses, /v1/messages
- Smallest pack
- $99.00
Latency and error rate per model are on the status page. status.cent-ai.dev
What a request costs
Open the full calculatorCost per request
Estimate
claude-sonnet-5.5An estimate at price version 6. Your bill comes from the usage the provider reports for each request.
Save $0.249Code
curl https://api.cent-ai.dev/v1/chat/completions \
-H "Authorization: Bearer $CENT_AI_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "claude-sonnet-5.5", "messages": [{"role": "user", "content": "Hello"}]}'Set CENT_AI_KEY to a key from the dashboard. OpenAI SDKs use https://api.cent-ai.dev/v1; Anthropic SDKs use https://api.cent-ai.dev.
Price history
Version 6, in effect since Oct 31, 2026.
Effective
per 1M tokens
Inputold $0.58 new $0.58Cache readold $0.058 new $0.058Cache writeold $0.725 new $0.58Cache write 1hold $1.16 new $0.58Outputold $2.90 new $2.90Effective
per 1M tokens
Inputold $0.30 new $0.58Cache readold $0.03 new $0.058Cache writeold $0.375 new $0.725Cache write 1hold $0.60 new $1.16Outputold $1.50 new $2.90Effective
per 1M tokens
Inputold $1.14 new $0.30Cache readold $0.114 new $0.03Cache writeold $1.425 new $0.375Cache write 1hold $2.28 new $0.60Outputold $5.70 new $1.50Effective
per 1M tokens
Inputold $1.14 new $1.14Cache readold $0.057 new $0.114Cache writeold $1.425 new $1.425Cache write 1hold $2.28 new $2.28Outputold $5.70 new $5.70Effective
per 1M tokens
Inputold $0.38 new $1.14Cache readold $0.019 new $0.057Cache writeold $0.475 new $1.425Cache write 1hold $0.76 new $2.28Outputold $1.90 new $5.70
Compare with
- deepseek-v4.1-flash-uncensoredInput$0.156
- gpt-6-astraInput$3.20
- claude-opus-5Input$1.15
- claude-fable-5Input$2.80
Questions
How much does claude-sonnet-5.5 cost on Cent AI?
$0.58 per 1M input tokens and $2.90 per 1M output tokens. Cache reads cost $0.058 and cache writes $0.58 per 1M tokens. Each request is rounded up once.
Is there a minimum charge?
A successful request costs at least $0.00002. A request that ends early, because the client left or the output was cut off, pays only for its tokens.
How do I pay?
Top up a prepaid USD balance with QRIS or USDT/USDC, or buy a budget pack once. Nothing renews automatically.
What if the provider does not report usage?
The request is billed from an estimate and marked estimated on the Requests page. An estimate never contains cache tokens.
Which endpoint do I call?
/v1/chat/completions, /v1/responses, /v1/messages. The OpenAI base URL is https://api.cent-ai.dev/v1.
Try claude-sonnet-5.5 with one key
One prepaid balance for every model. Pay with QRIS or USDT/USDC.