8 image models · capabilities and pricing in one placeSee pricing →
Models & pricing

TEXT MODEL DETAILS

Claude Opus 5.5

The same model can expose multiple Token-priced groups. Review input, output and cache prices before checking whether invocation is open.

Billing unit
1M tokens
Status
Selectable

TEXT MODEL · TOKEN PRICING

claude-opus-5-5 · billed per million tokens

Choose a model and pricing group. Input, output and cache are billed separately, with rates locked at task creation.

Selectable

Groups are separately priced options for the same model. Compare input, output and cache rates, then choose a group allowed by your API key. Without a group, the key allow list and routing strategy apply; there is no fixed default group. No unpublished latency or SLA is implied.

PRICING GROUP

Group 001

claude-opus-5-5

Available
Input / 1M
≈ $0.6618
Output / 1M
≈ $3.3091
Cache read / 1M
≈ $0.0662
Cache write / 1M
$0.827283
PRICING GROUP

Group 002

claude-opus-5-5

Available
Input / 1M
$5
Output / 1M
$25
Cache read / 1M
$5 (same as input; no cache-read discount)
Cache write / 1M
$5.000000
PRICING GROUP

Group 003

claude-opus-5-5

Available
Input / 1M
$8.5
Output / 1M
$42.5
Cache read / 1M
$8.5 (same as input; no cache-read discount)
Cache write / 1M
$8.500000
PRICING GROUP

Group 004

claude-opus-5-5

Available
Input / 1M
$36.5
Output / 1M
$182.5
Cache read / 1M
$36.5 (same as input; no cache-read discount)
Cache write / 1M
$36.500000
Compare input and output rates in a chart

Compare group rates

Compare input and output on the same unit; cache rates are listed above. Tasks settle with locked rates and verified usage.

USD / 1M Token
GroupInputOutput
claude-opus-5-5 · Group 0010.6618273.309132
claude-opus-5-5 · Group 0025.00000025.000000
claude-opus-5-5 · Group 0038.50000042.500000
claude-opus-5-5 · Group 00436.500000182.500000

Price snapshot: 2026-11-12 04:27 (UTC+8)

Invocation boundaries

  • Use the OGA-native text-task API or the OpenAI-compatible /v1/chat/completions endpoint with common open-source frameworks.
  • OpenAI-compatible clients use claude-opus-5-5 or one of its four group aliases; upstream URLs, supplier groups and credentials are never client inputs.
  • The concrete rate is locked at task creation; settlement uses final valid provider usage.
  • Unknown acceptance, missing usage or incomplete rates never trigger an automatic retry, group switch or free settlement.
Read Token integration notes