GitHub Copilot to Retire Six Models on Dec 14, With No Successor Yet for Kimi K3
GitHub said on Nov 13 that Claude Fable 5, Claude Haiku 4.5, Gemini 3.8 Flash, Gemini 4 Argon, Grok 4.6 and Kimi K3 will leave Copilot experiences on Dec 14.

What happened
The deprecation covers all Copilot experiences: Copilot Chat, inline edits, ask and agent modes, and code completions. GitHub’s changelog lists the six models with a suggested alternative for each: Fable 5 by Fable 5.2, Haiku 4.5 by Haiku 5.5, Gemini 3.8 Flash by Gemini 4 Hydrogen, Gemini 4 Argon by Gemini 4 Carbon, and Grok 4.6 by Grok 4.7; the alternative for Kimi K3 reads “to be announced.”
There is one exception: individual Copilot subscribers on annual plans will keep access to Claude Haiku 4.5. Under default model enablement, the suggested alternatives are automatically enabled for Copilot Enterprise and Copilot Business customers unless an administrator has turned off the global default or explicitly disabled the model; no action is required to remove the models after they have been deprecated.
This is the fourth deprecation wave in ten weeks: four models on Oct 2, six on Oct 19, and three more — Claude Opus 4.8, Claude Opus 5 and GPT-5.6 Terra — due on Nov 16, for 19 models retired or scheduled across the four waves.
What it means for users
For users, swapping models is not a simple rename: different models differ in answer quality, speed and price. GitHub says nothing about price in the changelog; on GitHub’s own rate card, some replacements cost more than the models they replace — Gemini 4 Carbon, for example, is set to nearly double in price on Dec 1.
Two models previously recommended as replacements are themselves on this wave’s list: Gemini 3.8 Flash and Kimi K3 were named as alternatives in October’s waves and retire on Dec 14 as well; Grok 4.6 was the successor to Grok 4.5, retired on Oct 19.
Key facts
- Retirement: On Dec 14, 2026, six models leave all Copilot experiences, including Chat, inline edits, ask and agent modes, and code completions
- List: Claude Fable 5, Claude Haiku 4.5, Gemini 3.8 Flash, Gemini 4 Argon, Grok 4.6 and Kimi K3
- Replacements: Fable 5 to Fable 5.2; Haiku 4.5 to Haiku 5.5; Gemini 3.8 Flash to Gemini 4 Hydrogen; Gemini 4 Argon to Gemini 4 Carbon; Grok 4.6 to Grok 4.7; Kimi K3 to be announced
- Exception: Individual subscribers on annual plans keep access to Claude Haiku 4.5
- Auto-enable: Under default settings, replacements are automatically enabled for Enterprise and Business customers; no manual removal is needed after deprecation
- Waves: The fourth wave in ten weeks; 19 models retired or scheduled across the four waves
Why it matters
Model churn has gone from occasional housekeeping to routine, with 19 models retired or scheduled in four waves over ten weeks; Kimi K3 users still have no officially named successor, and replacements differ in price and answer quality.
CompaniesAnthropic·Google·Microsoft·xAI · Musk·Moonshot AI
Grab todayOpenAI Codex for Students: $100 in Free CreditsHIMIAO Picks →Follow HIMIAO Horizon
No sign-up · Unsubscribe any timeNotifications
A ping on your phone or computer when the brief is out. No email needed.
Comments
Related
- Oct 8Google Launches Universal Work Agent Gemini Agent, Running Gemini and Anthropic's Claude
- Oct 18GitHub Copilot to Remove Three Flagship Models on November 16: Claude Opus 4.8/5 and GPT-5.6 Terra Retired
- Nov 11Anthropic Gives Claude an Office Suite: Meetings, Sheets and Whiteboards
- Oct 9UK's ICO Names 10 AI Labs in Training-Data Privacy Crackdown: "No Justification" for Non-Compliance
- Oct 22Claude Code vs Codex CLI vs Gemini CLI: Three Terminal Coding Agents Compared
Today
Nov 15 Sunday- Brief
- GitHubGitHub Weekly Vol.5: 6 Projects Worth a Look This Week
Tested or read — weeklyAll →
- Quiz
- Call
Latest news
All →- Nov 15Perplexity Open-Sources Decider 1.3, Adding Text-Span Extraction
- Nov 15Microsoft 365 Outage Lasts Over 24 Hours After Network Upgrade Fault
- Nov 15Fed Report: AI Becomes the Top Financial-Stability Risk for the First Time
- Nov 15FBI Warns Hackers Have Weaponized the ARTEX AI Pen-Testing Framework
- Nov 15Applied Materials Posts Record Q4 Revenue of $10.41B, Up 53%
- Nov 15Anthropic IPO Said to Turn Away China, Hong Kong Orders
™