MiniMax
MediumVerified
Developer API · Jun 20, 2025
Text-to-voice API becomes legacy
The older API remains operational but is no longer updated.
Public preview · Automatic monitoring and email alerts are not enabled
Build snapshot for UTC 2026-09-03, not live metrics. JavaScript recalculates date states for the current UTC day; sources are not reverified.
Normalized event timeline
Separate what is new from what needs action.
Filter across providers, event types, severity, deadline and verification state. Every verified record links to a public primary source.
Showing all 20 events
No deadline statedNot announced
Cohere
HighVerified
Developer API · Not announced
Embedding v2 models retired
The listed embedding models retired on Apr 4, 2026.
Mistral
MediumVerified
Developer API · Not announced
mistral-medium-2508 is deprecated
The model card marks mistral-medium-2508 deprecated and recommends Mistral Medium 3.5.
MiniMax
LowVerified
Developer API · Jun 1, 2026
MiniMax-M3 is documented for the API
MiniMax-M3 is listed in both release notes and the international API model catalog.
DeepSeek
HighVerified
Developer API · Apr 24, 2026
Legacy aliases retired: deepseek-chat, deepseek-reasoner
The listed aliases were scheduled for discontinuation on Jul 24, 2026. Replacement: deepseek-v4-flash.
xAI
MediumVerified
Model alias routing · Jul 29, 2026
grok-voice-latest routes to grok-voice-think-fast-2.0
xAI announced that the grok-voice-latest alias would route to grok-voice-think-fast-2.0 starting Aug 5, 2026.
MiniMax
HighVerified
Developer API · Not announced
Music API access changes
Music API access changed: Paid APIs closed to new users; existing paying users continue. Free APIs to be discontinued.
DeepSeek
LowVerified
Developer API · Aug 21, 2026
deepseek-v4-flash-vision-exp is documented for the API
deepseek-v4-flash-vision-exp is available as an experimental API model.
Z.ai / GLM
LowVerified
Developer API · Aug 26, 2026
glm-5.3-flash is documented for the API
glm-5.3-flash was announced on Aug 26, 2026; its model page separately documents Model API access.
xAI
HighVerified
Request defaults and billing · Aug 28, 2026
Image quality default changed for grok-imagine-image-2.0
xAI changed the omitted quality behavior for grok-imagine-image-2.0: Omitted quality defaults to auto; generation currently resolves to low and editing to medium.
OpenAI
CriticalVerified
API and model lifecycle · Mar 24, 2026
Listed video models and API scheduled for shutdown
Requests to the listed models and API are scheduled to stop working on Sep 24, 2026.
OpenAI
CriticalVerified
Model lifecycle · Sep 26, 2025
4 legacy completion models reach shutdown
OpenAI lists Sep 28, 2026 as the shutdown date for 4 legacy model identifiers.
OpenAI
CriticalVerified
Model lifecycle · Apr 22, 2026
Legacy GPT, reasoning and image snapshots scheduled for shutdown
A broad OpenAI model shutdown wave is listed for Oct 23, 2026, with substitutes varying by workload.
OpenAI
HighVerified
Platform capability · May 7, 2026
Self-serve fine-tuning job creation closes for active existing customers
OpenAI states that active existing customers will no longer be able to create new self-serve fine-tuning jobs from Jan 6, 2027.
Google Gemini
MediumVerified
Model lifecycle · Not announced
gemini-3.1-flash-lite has a stated shutdown window
Google's deprecation table lists Earliest: May 7, 2027 for gemini-3.1-flash-lite and recommends gemini-3.5-flash-lite.
Google Gemini
LowVerified
Model lifecycle · Not announced
gemini-embedding-001 receives a shutdown date
Google lists May 14, 2028 as the shutdown date for gemini-embedding-001 and recommends gemini-embedding-2.
Anthropic
HighVerified
Model lifecycle · Not announced
claude-mythos-preview marked deprecated
Anthropic's model status page marks claude-mythos-preview as deprecated and points developers to claude-mythos-5.
Anthropic
HighVerified
Request parameters · Not announced
Sampling parameters deprecated on Claude Opus 4.7 and later
Anthropic documents the following behavior on the affected models: Omit parameters; non-default values return HTTP 400.
Alibaba / Qwen
HighNeeds Review
Developer API · Not announced
Bailian retirement notices need model and date reconciliation
The retirement index and linked notice header disagree; affected model IDs and regions could not be confirmed.
Moonshot / Kimi
LowVerified
Developer API · Not announced
kimi-k3 is documented for the API
The current international API quickstart uses kimi-k3; its launch date is not stated.
No events match these filters.
Broaden the date range or reset one of the provider, severity or verification filters.