14 changes verified September 11, 2026

Know what will break before it ships.

A source-backed compatibility database for AI APIs. Find model shutdowns, exact error symptoms, affected identifiers, and the migration fix before production finds them for you.

Official sources monitored by cron
Exact model and parameter IDs
Visible verification dates

Stack compatibility planLocal scan

Paste model IDs, API parameters, endpoints, or code

Your input stays in this browser. Nothing is uploaded or stored.

14Verified changes

12Platforms monitored

4Upcoming cutoffs

100%Official-source records

Deadline watch

Cutoffs that need action

View deadline calendar →

17 days leftSeptember 28, 2026

Four legacy OpenAI models shut down on September 28, 2026

Choose gpt-5.4-mini or gpt-5-mini based on workload requirements, then migrate and test the integration.

OpenAI · OpenAI API 19 days leftSeptember 30, 2026

labs-leanstral-1-5: model retirement on September 30, 2026

No replacement identifier passed the evidence gate. Follow the linked provider documentation for the required request or behavior change.

Mistral · Mistral API 40 days leftOctober 21, 2026

o1: model retirement on October 21, 2026

The official source lists gpt-5.6-sol as a replacement identifier. Test the migration against current provider documentation before production.

Microsoft · Azure AI Foundry 76 days leftNovember 26, 2026

Vertex AI Extensions: api sunset on November 26, 2026

The official source lists Agent Platform as a replacement identifier. Test the migration against current provider documentation before production.

Google · Vertex AI

Change ledger

Recently announced or effective

Browse all changes →

MicrosoftAzure AI Foundryhigh

o1: model retirement on October 21, 2026

Microsoft documents a model retirement affecting o1 on October 21, 2026. The source lists gpt-5.6-sol as a replacement identifier.

model retirementSearch repositories and deployment configuration for o1. The official source lists gpt-5.6-sol as a replacement identifier. Test the migration against current provider documentation before production.

MistralMistral APImedium

labs-leanstral-1-5: model retirement on September 30, 2026

Mistral documents a model retirement affecting labs-leanstral-1-5 on September 30, 2026. No replacement identifier passed the evidence gate.

model retirementSearch repositories and deployment configuration for labs-leanstral-1-5. No replacement identifier passed the evidence gate. Follow the linked provider documentation for the required request or behavior change.

MicrosoftAzure AI Foundryhigh

TimeGEN-1: model retirement on August 31, 2026

Microsoft documents a model retirement affecting TimeGEN-1 on August 31, 2026. The source lists TimeGPT-1 and TimegGPT-2.1 as replacement identifiers.

model retirementSearch repositories and deployment configuration for TimeGEN-1. The official source lists TimeGPT-1 and TimegGPT-2.1 as replacement identifiers. Test the migration against current provider documentation before production.

GoogleGemini APIhigh

Gemini deprecates temperature, top_p, and top_k

Gemini 3.6 Flash and 3.5 Flash-Lite ignore the sampling parameters; Google says future model generations will return HTTP 400 when they are supplied.

parameter deprecationStrip temperature, top_p, and top_k from Gemini requests and encode response constraints in a clear system instruction.

AnthropicClaude APIhigh

/v1/experimental/generate_prompt, /v1/experimental/improve_prompt, and 1 more: api sunset on August 17, 2026

Anthropic documents a api sunset affecting /v1/experimental/generate_prompt, /v1/experimental/improve_prompt, and 1 more on August 17, 2026. No replacement identifier passed the evidence gate.

api sunsetSearch repositories and deployment configuration for /v1/experimental/generate_prompt, /v1/experimental/improve_prompt, and /v1/experimental/templatize_prompt. No replacement identifier passed the evidence gate. Follow the linked provider documentation for the required request or behavior change.

AnthropicClaude APIhigh

Claude Sonnet 5 changes thinking defaults and token counts

Claude Sonnet 5 enables adaptive thinking by default, removes manual thinking budgets, rejects non-default sampling parameters, and tokenizes the same text into roughly 30% more tokens.

breaking behaviorRemove manual thinking budgets and non-default sampling parameters, rely on adaptive thinking, and use Anthropic's token-counting API to recalibrate limits and cost controls.

From symptom to fix

A changelog that answers the next question.

Provider announcements tell you what changed. Each record here also maps the affected surface, the observable failure, and the practical migration step.

1Paste a model ID, parameter, code fragment, or error.
2Match it against source-backed provider announcements.
3Open the source receipt and apply the migration fix.

anthropic / request.pyHTTP 400 → valid request

response = client.messages.create(
    model="claude-opus-4-8",
-   temperature=0.2,
-   top_p=0.9,
+   system="Be concise and factual.",
    messages=messages,
)

Coverage