Model retirements this month

The Model retirements this month is 3.00 on 2026-10-23 (n=3 sources). Definition: Models retired in the calendar month: first-party scheduled shutdowns (provider deprecation pages) plus models removed by hosts (DeepInfra deprecated flag flips and delistings, observed). The two components are published separately in the metadata.

Computed from the SKU state at the end of each UTC day (latest value: 2026-10-23; today's value is recomputed until the day ends). The constituents of the latest value are listed below; each shows when it was verified and links to its source.

Constituents of the latest value

Details of the latest value

month

2026-10

litellm deprecation date crosscheck

93

gap

no first-party deprecation-page recipe yet (openai/anthropic/google-deprecations carry no fields): the value counts DeepInfra host-level retirements only; LiteLLM deprecation dates are shown as a cross-check, not added

{
 "components": {
  "removed_by_hosts": 3,
  "scheduled_first_party": 0
 },
 "host_delisted_in_month": [
  "llm/deepinfra/openai/gpt-oss-120b-turbo",
  "llm/deepinfra/thinkingmachines/inkling"
 ],
 "host_scheduled_in_month": [
  "llm/deepinfra/gryphe/mythomax-l2-13b"
 ],
 "removed_by_hosts_sku_ids": [
  "llm/deepinfra/gryphe/mythomax-l2-13b",
  "llm/deepinfra/openai/gpt-oss-120b-turbo",
  "llm/deepinfra/thinkingmachines/inkling"
 ]
}

Definition

Models retired in the calendar month: first-party scheduled shutdowns (provider deprecation pages) plus models removed by hosts (DeepInfra deprecated flag flips and delistings, observed). The two components are published separately in the metadata.

Inclusion rule

scheduled: SKUs from the openai/anthropic/google deprecation-page recipes whose stated shutdown date is in the month (none until those recipes exist); host level: DeepInfra models whose stated shutdown timestamp (deprecated_at) is in the month, plus DeepInfra models delisted in the month (observed), each model once. The two components are in the metadata. LiteLLM deprecation_date is a cross-check, not added.

When it cannot be computed: Not computed: that day shows no number and states the reason; the last full value stays shown with its date. A source whose last confirmed snapshot is older than 3x its cadence (6h sources 18 h, 12h sources 36 h, 1d sources 72 h) drops out; below minimum coverage the series is not computed and says 'not verified: <source> stale since <date>'.

Minimum coverage: 1 sources. Below that the series is not computed and says so instead of showing a number.

Sources feeding it

openai-deprecations, anthropic-deprecations, google-deprecations, deepinfra-models

Deliberately excluded

• litellm deprecation dates (shown as cross-check)

What can break it

• providers describe retirements in prose; one notice can cover several models
• dates move (a new dated fact)
• no probing of provider /models endpoints
• a DeepInfra model that disappears for non-retirement reasons is counted

Daily values, last 90 days

Every day since 2023-09-07: CSV · JSON.

API: JSON · CSV

Rack Rates (2026). Model retirements this month [data as of 2026-10-23]. https://rackrates.dev/index/model-retirements-month (accessed 2026-10-23). List prices, not transacted prices. CC BY 4.0.

@misc{rackrates_index_model_retirements_month_2026,
  author       = {{Rack Rates}},
  title        = {Model retirements this month},
  year         = {2026},
  howpublished = {\url{https://rackrates.dev/index/model-retirements-month}},
  note         = {Data as of 2026-10-23; accessed 2026-10-23; list prices, not transacted prices; CC BY 4.0}
}