AI model shutdown and price change dates
Every announced model shutdown, price change and promotion end we track, with the model to move to and a link to the provider’s notice. Next up, 21 November 2026: GPT-5.6 Sol's promotional $4 / $20 price is guaranteed until at least this date.
Add every date to your calendar: Calendar feed (.ics) · Past changes: Price history
November 2026
1 date
Promotion may end GPT-5.6 Sol's promotional $4 / $20 price is guaranteed until at least this date. (GPT-5.6 Sol) Source
December 2026
4 dates
Price change Gemini 4 Argon's introductory price ends. From today it costs $4 / $20 per 1M tokens, with cache reads at $0.20. (Gemini 4 Argon) Source- Shutdown Requests to Gemini Omni 1.1 Flash start running on Gemini Omni 1.2 Flash and are billed at 1.2's lower rates: about $0.07 a second of 720p video instead of $0.10. Source
Shutdown Moonshot plans to have retired all Kimi K2 models by the end of December. (Kimi K2.7 Code, Kimi K2.6) Source Move to Kimi K3 ($3 / $15)
January 2027
2 dates
- Price change Gemini 3.8 Flash's promotional price ends. From today it costs $1.50 / $7.50 per 1M tokens, with cache reads at $0.15. (Gemini 3.8 Flash, Gemini 3.7 Flash, Gemini 3.6 Flash) Source
- Shutdown OpenAI shuts down its text-to-speech models tts-1, tts-1-hd and gpt-4o-mini-tts. It recommends moving to gpt-realtime-2.1-mini through the Realtime API. Source
March 2027
1 date
- Shutdown Google shuts down Nano Banana (gemini-2.5-flash-image) and recommends moving to Nano Banana 2 Lite, which costs $0.0336 an image at 1K. (Nano Banana) Source
April 2027
1 date
Shutdown Mistral retires Large 3 and Medium 3.5. It recommends moving to Mistral Large 4. (Mistral Large 3, Mistral Medium 3.5) Source Move to Mistral Large 4 ($0.70 / $2.10)
May 2027
1 date
- Shutdown Google shuts down Gemini 3.1 Flash-Lite. It recommends moving to Gemini 3.5 Flash-Lite. (Gemini 3.1 Flash-Lite) Source Move to Gemini 3.5 Flash-Lite ($0.30 / $2.50)
Models being retired
6 models
| Model | Shuts down | Price now | Move to |
|---|---|---|---|
| o3OpenAI | $2.00 / $8.00 | GPT-6 Sol$2 / $10 | |
| Kimi K2.7 CodeMoonshot AI | $0.95 / $4.00 | Not named | |
| Kimi K2.6Moonshot AI | $0.95 / $4.00 | Kimi K3$3 / $15 | |
| Mistral Medium 3.5Mistral AI | $1.50 / $7.50 | Mistral Large 4$0.70 / $2.10 | |
| Mistral Large 3Mistral AI | $0.50 / $1.50 | Mistral Large 4$0.70 / $2.10 | |
| Gemini 3.1 Flash-LiteGoogle | $0.25 / $1.50 | Gemini 3.5 Flash-Lite$0.30 / $2.50 |
Prices per 1M input / output tokens. The replacement is the one the provider recommends; check its price and quality on your own work before the date.
Promotional prices and when they end
| Model | Until | Price now | Afterwards |
|---|---|---|---|
| GPT-5.6 SolOpenAI | at least | $4.00 / $20.00 | Not announced |
| Gemini 4 ArgonGoogle | $2.00 / $10.00 | $4.00 / $20.00 | |
| Gemini 3.8 FlashGoogle | $0.75 / $3.75 | $1.50 / $7.50 | |
| Gemini 3.7 FlashGoogle | $0.75 / $3.75 | $1.50 / $7.50 | |
| Gemini 3.6 FlashGoogle | $0.75 / $3.75 | $1.50 / $7.50 |
Per 1M input / output tokens. “At least” means the provider guarantees the price until that date but hasn’t said what happens after.
Recently retired or repriced
Full price history- Mistral deprecated Large 3 and Medium 3.5, which retire on 30 April 2027, and pointed users to Mistral Large 4, now listed at $0.70 / $2.10 per 1M tokens (it was $1.36 / $4.18 when we checked on 22 October). (Mistral Large 4, Mistral Large 3, Mistral Medium 3.5) Source
- The cheapest OpenRouter hosts repriced 6 open-weight models, mostly downwards, including gpt-oss-120b (now $0.03 / $0.15) and Qwen3.8 27B (now $0.039 / $1.25). (gpt-oss-120b, Gemma 4 26B-A4B, Qwen3.8 27B, Qwen3.6 35B-A3B, Nemotron 3 Super, Nemotron 3.5 Lightning) Source
Anthropic halved Claude Sonnet 5.5's cache-read price from $0.20 to $0.10 per 1M tokens; Anthropic says this cuts most agent workloads by about 20%. (Claude Sonnet 5.5) Source
DeepSeek V4 Pro moved to peak and off-peak pricing: $1.32 / $3.96 at peak, down from $1.74 on input and up from $3.48 on output.$1.74 / $3.48→ $1.32 / $3.96 (DeepSeek V4 Pro) Source- Anthropic retired Claude Opus 4.1. (Claude Opus 4.1) Source
- Mistral retired Nemo, Small 3.2, Devstral 2 and Magistral 1.2. (Mistral Nemo, Mistral Small 3.2) Source
Questions
What is the next AI pricing or model date to know?
21 November 2026: GPT-5.6 Sol's promotional $4 / $20 price is guaranteed until at least this date.
Which AI models are being shut down?
Gemini Omni 1.1 Flash on 3 Dec 2026, o3 on 11 Dec 2026, Kimi K2.7 Code and Kimi K2.6 on 31 Dec 2026, OpenAI tts-1, tts-1-hd and gpt-4o-mini-tts on 6 Jan 2027, Nano Banana on 15 Mar 2027, Mistral Large 3 and Mistral Medium 3.5 on 30 Apr 2027 and Gemini 3.1 Flash-Lite on 7 May 2027. Each entry above links to the provider’s notice and the model it recommends instead.
When does Gemini 4 Argon’s price change?
On 1 December 2026. Gemini 4 Argon's introductory price ends. From today it costs $4 / $20 per 1M tokens, with cache reads at $0.20.
When does Gemini 3.8 Flash’s price change?
On 1 January 2027. Gemini 3.8 Flash's promotional price ends. From today it costs $1.50 / $7.50 per 1M tokens, with cache reads at $0.15.
What happens when a model is shut down?
Requests to the retired model ID stop working, so apps that still call it start returning errors. Switch to the replacement and test it before the date; replacements often cost a different amount per token.
How can I keep track of these dates?
Add the calendar feed (https://aitokenprice.com/dates.ics) to Google Calendar, Apple Calendar or Outlook as a subscription, and new dates appear on their own. Price changes that have happened are logged on the price history page and in its RSS feed.