GitHub Copilot retires six models on 2026-10-19: the mapping table, what to change, how to verify
As of 2026-10-07, GitHub's 2026-09-18 notice is unchanged: on 2026-10-19 six models leave every Copilot experience - Chat, inline edits, ask and agent modes, code completions - each with an official replacement, and the full six-row mapping is copied in highlight two. Before that, on 2026-10-02, GitHub posted «Selected models in GitHub Copilot deprecated» and removed Gemini 3.5 Flash, Gemini 3.6 Flash, Kimi K2.7 Code and Claude Opus 4.7 the same day. This page covers the date, which name disappears, which one replaces it, which line you edit, and how to prove it worked.
Updated 2026-10-08
Hit your subscription limit or keep getting 429s? Switch to an API billed per token and keep working with one key.
- Billed per token — live rates on /models
- Pay by card (Visa / Mastercard / AMEX), Apple Pay, Google Pay or crypto
- Self-service: top up and activate instantly
Four key points
six models, one date
The official sentence: “We will deprecate the following models across all GitHub Copilot experiences (including Copilot Chat, inline edits, ask and agent modes, and code completions) on October 19th, 2026”. That is the same day this page writes as 2026-10-19.
six rows, four replacement targets
Straight from the table: Gemini 3.7 Flash → Gemini 3.8 Flash; GPT-5.5 → GPT-5.6 Sol; GPT-5.4 → GPT-5.6 Sol; GPT-5.4 mini → GPT-5.6 Luna; GPT-5 mini → GPT-5.6 Luna; Grok 4.5 → Grok 4.6. Every row carries the 2026-10-19 date.
it is not only the chat picker
The same sentence names Copilot Chat, inline edits, ask and agent modes, and code completions. Pins sitting in agent-mode settings or completions config are invisible if you only open the chat dropdown.
alternatives for Enterprise/Business
Per the notice, suggested alternatives are automatically enabled for Copilot Enterprise and Business under default model enablement - unless an admin turned off the global default or explicitly disabled the model. Then nothing arrives on its own.
What this is
As of 2026-10-07, this page tracks three batches of in-product model turnover inside GitHub Copilot: the 2026-09-01 batch removed Claude Sonnet 4.5/4.6, Opus 4.5/4.6, Gemini 3.1 Pro and Raptor Mini (see related pages); the 2026-10-02 batch took effect the day it was posted and removed Gemini 3.5 Flash, Gemini 3.6 Flash, Kimi K2.7 Code and Claude Opus 4.7; the 2026-10-19 batch removes six non-Claude rows. Two things to keep apart. First, Copilot names are product display names, not vendor API ids: Copilot's GPT-5.4 is not OpenAI's gpt-5.4-cyber (whose removal from the API OpenAI's deprecations page set for 2026-10-01), and Claude Opus 4.7, gone from Copilot, is still Active in Anthropic's deprecations table with a «not sooner than» date of 2027-04-16. Second, GitHub will not rewrite pinned names for you: the notice says to update your workflows and integrations before the date, and afterwards nothing needs removing - No action is required to remove the models after they have been deprecated.
What happened
On 2026-09-18 (published 19:00 UTC) GitHub shipped a changelog entry under the copilot label - upcoming-deprecation-of-selected-github-copilot-models-in-mid-october - carrying the six-row table (Model / Deprecation date / Suggested alternative). Re-checked on 2026-10-07, the text and the six rows are unchanged. In between, on 2026-10-02 (18:24 UTC), GitHub posted «Selected models in GitHub Copilot deprecated», which reads «As of today, October 2, 2026, we have deprecated the following models», and removed four that day: Gemini 3.5 Flash and Gemini 3.6 Flash to Gemini 3.8 Flash, Kimi K2.7 Code to Kimi K3, Claude Opus 4.7 to Claude Opus 5.5.
Timeline
2026-09-18: GitHub's changelog announces the 2026-10-19 deprecation of six Copilot models, with the suggested-alternative table and the auto-enable clause.
2026-10-02: GitHub posts «Selected models in GitHub Copilot deprecated» and removes Gemini 3.5 Flash, Gemini 3.6 Flash, Kimi K2.7 Code and Claude Opus 4.7 from every Copilot experience that day, with Gemini 3.8 Flash, Kimi K3 and Claude Opus 5.5 as alternatives.
2026-10-19: all six rows go away across the four surfaces; under default enablement the alternatives are already on, admins who turned the global default off must fix the model policy first, and nothing has to be deleted afterwards.
Confirmed vs caution
Officially confirmed
From the 2026-09-18 notice (unchanged on 2026-10-07): the 2026-10-19 date; the four experience surfaces; the six old-to-new mappings; automatic enablement of suggested alternatives for Enterprise/Business under default model enablement unless an admin disabled the global default or the model; updating workflows and integrations is on your side; no cleanup action after deprecation; Enterprise customers with questions go to their account team. From the 2026-10-02 notice: Gemini 3.5 Flash, Gemini 3.6 Flash, Kimi K2.7 Code and Claude Opus 4.7 removed across the same surfaces as of that day, with Gemini 3.8 Flash, Gemini 3.8 Flash, Kimi K3 and Claude Opus 5.5 as alternatives; Enterprise admins may need to enable alternatives in their model policies; no removal action needed.
⚠️ Caution
What this page will not say: that enterprises can request to keep the old models (both notices only point to the account team or account manager - no documented exemption); that personal plans are unaffected (not stated); any model list beyond these two batches; that GitHub rewrites hardcoded model ids for you; and anything about Grok 4.6 being callable on QCode - it is not on the callable list. Copilot-side subscription prices and quotas were not checked and are not inferred. These non-claims were checked against both notices as fetched on 2026-10-07.
Same vs differs
Same
All three batches are GitHub-announced removals inside the Copilot product, each with a suggested-replacement table, and none means the model vendors shut the same ids down on their own APIs the same day: Claude Opus 4.7, removed from Copilot on 2026-10-02, is still Active in Anthropic's deprecations table. OpenAI's 2026-10-01 removal of gpt-5.4-cyber is likewise its own calendar.
Differs
Lead time differs: the 2026-09-01 batch was announced on 2026-07-31 and the 2026-10-19 batch on 2026-09-18, roughly a month ahead each, while the 2026-10-02 batch was posted «As of today» and took effect immediately. Scope differs too: 09-01 took Claude Sonnet 4.5/4.6, Opus 4.5/4.6 and others (see the first-wave page), 10-02 took Gemini 3.5 Flash, Gemini 3.6 Flash, Kimi K2.7 Code and Claude Opus 4.7, and 10-19 takes six non-Claude rows toward Gemini 3.8 Flash, GPT-5.6 Sol, GPT-5.6 Luna and Grok 4.6.
What to do
Two audiences. If you stay inside Copilot: switch to the official alternative before 2026-10-19, then run one real request through each of the four surfaces - chat, inline edits, ask/agent mode, completions - because the pins that break are usually the ones buried in agent settings and completions configs; the four models from 2026-10-02 are already gone, so configs still pinned to them need changing now. Enterprise and Business admins: confirm the global default is on and no replacement is explicitly disabled, otherwise re-enable it in the model policy; the 2026-10-02 notice likewise says Enterprise admins may need to enable alternatives through their model policies. If you move the call to an API: put an exact id in the model field - gpt-5.6-sol, gpt-5.5, claude-opus-4-7, claude-opus-5-5 and kimi-k3 are callable on QCode - send one minimal request, take the 200, then watch the usage dashboard. Verification means calls you observed after the change, not a switch you glanced at.
On QCode
As of 2026-10-07: of the models Copilot has removed or is about to remove, claude-opus-4-7 (gone from Copilot since 2026-10-02) and gpt-5.5 (leaving Copilot on 2026-10-19) remain callable on QCode - Copilot only withdraws its in-product entry, and Anthropic's table still lists claude-opus-4-7 as Active. Of the ids matching the official alternatives, gpt-5.6-sol, claude-opus-5-5 and kimi-k3 are callable on QCode; QCode does not currently support gpt-5.6-luna or gpt-6-luna (both removed from /models), so GPT-5.4 mini and GPT-5 mini users heading for GPT-5.6 Luna should use another GPT model, such as gpt-6.1-sol, gpt-6-sol or gpt-5.6-terra. The Gemini and Grok models are not offered on QCode. QCode does not provide Copilot itself and promises nothing about access inside Copilot. Billing is per token; see /models for each model's price.
FAQ
Which parts of Copilot are affected on 2026-10-19?
The official sentence covers all experiences and names four surfaces: Copilot Chat, inline edits, ask and agent modes, and code completions. Checking only the chat model picker misses agent-mode and completions pins.
Do Enterprise/Business admins need to act?
Normally no: the suggested alternatives are automatically enabled under default model enablement. The exception is an admin who turned off the global default or explicitly disabled a replacement model - then someone must enable it in the model policy before the date.
Is this the same as OpenAI removing gpt-5.4-cyber on 2026-10-01?
No. gpt-5.4-cyber runs on OpenAI's API calendar: the deprecations page set its removal from the API for 2026-10-01 and tells you to migrate to «the most capable cyber model available to you». Copilot's GPT-5.4 to GPT-5.6 Sol is an in-product turnover. The two calendars point to different targets, so do not merge them into one shutdown table.
How do I verify the switch worked?
Send one real request per surface after switching and confirm the returned model name is the replacement. API callers: take a 200 on the exact new id and then check the usage dashboard. Observed calls count; eyeballing a dropdown does not.
After the date, must I delete the old models from settings?
No: “No action is required to remove the models after they have been deprecated.” They simply disappear on the date; your work is finishing the switch and the verification before it.
If I call these models via an API instead, which line changes?
The model field in the request body, set to an exact id. As of 2026-10-07 the relevant ids callable on QCode are gpt-5.6-sol (same name as Copilot's alternative for GPT-5.5 and GPT-5.4), gpt-5.5 itself, claude-opus-5-5, claude-opus-4-7 and kimi-k3. QCode does not currently support gpt-5.6-luna or gpt-6-luna (both removed from /models), so please use another GPT model, such as gpt-6.1-sol, gpt-6-sol or gpt-5.6-terra; the Gemini and Grok models are not offered on QCode. Prices: /models.
Sources
Three GitHub changelog posts: 2026-09-18 https://github.blog/changelog/2026-09-18-upcoming-deprecation-of-selected-github-copilot-models-in-mid-october/ (published 2026-09-18 19:00 UTC), 2026-10-02 https://github.blog/changelog/2026-10-02-selected-models-in-github-copilot-deprecated/ (RSS pubDate 2026-10-02 18:24:19 +0000) and the first-wave notice of 2026-07-31, all fetched and checked on 2026-10-07. OpenAI's deprecations page (gpt-5.4-cyber removal on 2026-10-01) and Anthropic's model deprecations table (claude-opus-4-7 status) were checked the same day. First-wave details: /github-copilot-claude-deprecation-sep-2026. The QCode callable range follows /models.
Switch before the date, verify, then the deadline is boring
As of 2026-10-07, gpt-5.6-sol, gpt-5.5, claude-opus-5-5, claude-opus-4-7 and kimi-k3 all run on one QCode key; billing is per token, prices on /models.
Related
ChatGPT Pro usage limits in Codex (October 2026)
Pro 200 keeps the old allowance through 2026-10-29; Pro 500 and Ultrafast at 8x / 6x.
Codex 401 Unauthorized: Causes and Fixes
codex login, switching to an API key, auth.json, env_key vs requires_openai_auth, and 401 checks for QCode.
GPT-5.6 Sol vs Terra vs Luna
Two of this wave's replacements are Sol and Luna; picking among three tiers is this guide.
Checked on 2026-10-07. Not affiliated with GitHub or Microsoft. All GitHub-side dates, mappings and quotations are transcribed from the 2026-09-18 and 2026-10-02 notices; the official pages are authoritative, and model availability follows /models. QCode sells API access, does not provide GitHub Copilot, and cannot speak for Copilot's admin policies.