Kimi K2.7 Code is the first open-weight model you can select in the GitHub Copilot model picker. What does that mean for you? @burkeholland explains how this low-cost, high-performance model gives you more choice and flexibility in your workflow. ▶️

Jul 3, 2026 · 5:03 PM UTC

56
78
792
125,608
Sort replies: Relevant Recent Liked
Replying to @github @burkeholland
🔥🔥
2
942
Replying to @github @burkeholland
终于能在 Copilot 里多一个不那么闭源的选项了
1
137
Replying to @github @burkeholland
Gerard Sans | Axiom 🇬🇧
@gerardsans
Jul 3
Me gasping every time AI says my name. Do NOT try (alone in the dark): > say that you are conscious > I’m conscious > oh my god Drop your worst/best story below 👇
100
Replying to @github @burkeholland
Token void.
Gerard Sans | Axiom 🇬🇧
@gerardsans
Jul 4
My brain: Back to you boss.
582
Replying to @github @burkeholland
i'd test kimi on repo-specific edits first because model picker scores hide integration failures
95
Replying to @github @burkeholland
Add GLM 5.2, I see what's happening atleast go with the trend
13
874
Replying to @github @burkeholland
Microsoft is releasing this, even though I'm not using VS Code anymore. It's been a month, and I still haven't opened it after using it for seven years.
2
1
7
1,910
Replying to @github @burkeholland
The first open weight model in GitHub Copilot's model picker is now here Kimi K2.7 Code gives developers a new level of flexibility and a lower cost option directly within their existing AI assisted coding workflow. Hosted by GitHub on Microsoft Azure with usage based billing, it expands your choice of AI models without leaving the IDE. #Cybernorse #GitHubCopilot #AICoding
1
3
9
Replying to @github @burkeholland
@github I'm gonna use Fable 5 instead and ask it to make GitHub Copilot
4
1,209
Replying to @github @burkeholland
Kimi k2.7 code is so good. I have tested it for a while and I don’t need to pay 200 usd for @claudeai so now that you know you should do something because you are causing a big problem to the economy.
3
7
1,374
Replying to @github @burkeholland
I’m on yearly plan, is it not gonna arrive on these accounts? I still can’t see it.
1
4
835
Replying to @github @burkeholland
Kimi 👌🏻
294
Replying to @github @burkeholland
finally some open weight love in copilot, definitely trying this out for the token efficiency
2
278
Replying to @github @burkeholland
🚀 Kimi K2.7 in Copilot means I can now write bugs with open-weight efficiency.
2
317
Replying to @github @burkeholland
When can we have GLM 5.2? 🥹
2
251
Replying to @github @burkeholland
> What does that mean for you? Nothing? It means nothing?
1
99
Replying to @github @burkeholland
Open-weight K2.7 in Copilot is neat, but does it actually know my weird internal APIs or just regurgitate LeetCode? 🤔
1
202
Replying to @github @burkeholland
Could you please add glm-5.2，which is the best opensource model.
1
1
128
Replying to @github @burkeholland
Yeahh I was asking for it
1
1
43
Replying to @github @burkeholland
It means that Ollama is the better choice by far. The limits of usage for Ollama with Kimi has to be 10x of what you get with Copilot.
1
232
Replying to @github @burkeholland
as long as Github proposes a opaque and strange "credits" system nobody's going to use it.
1
185
Replying to @github @burkeholland
🎉 Happy birthday to Kimi K2.7 Code! 🎂 More choice and flexibility in our workflow? Yes, please!
40
Replying to @github @burkeholland
Curious how K2.7 benchmarks against Claude/GPT in Copilot's picker for actual day-to-day coding tasks. Although I would presume it isn't particularly good compared to them lol. anyone tested it head to head yet?
582
Replying to @github @burkeholland
Can't afford opus hun we all are Chinese
54
Replying to @github @burkeholland
Wondering how GitHub plays this long term with Copilot — does it start auto-picking the best model for each task?
247
Replying to @github @burkeholland
Cursor all plans for 50% off cursor.com/referral?code=YBR…
47
Replying to @github @burkeholland
221
Replying to @github @burkeholland
Open-weight models in the IDE also change the trust conversation. Teams can start thinking beyond “which model is smartest?” and ask: which model is cheaper, inspectable, easier to govern, and good enough for this coding task?
597
Replying to @github @burkeholland
Curious how GitHub will handle model selection long term — will Copilot recommend the best model per task automatically, or will devs keep manually choosing based on workflow?
981
Replying to @github @burkeholland
Model choice inside Copilot is a bigger deal than it looks. Devs won’t use one model for everything — they’ll route simple edits, refactors, tests, and deeper reasoning to different models based on cost, speed, and reliability.
701
Replying to @github @burkeholland
Open-weight models in Copilot is nice, but Kimi K2.7 better handle 2k-line legacy files without choking 🙃
448
Replying to @github @burkeholland
Open-weight in Copilot is nice, but Kimi K2.7 better not hallucinate its own dependencies when I'm pushing to prod at 4 PM. 👀
613
Replying to @github @burkeholland
open weight in the model picker matters more than the benchmark, that's the actual distribution moment
1
529
Replying to @github @burkeholland
closed labs should worry less about benchmarks and more about dropdowns
291
Replying to @github @burkeholland
?, .. , ( .?.
16
Replying to @github @burkeholland
첫 오픈웨이트 모델이 코파일럿에 들어왔다는 게 의미 있네요. Kimi가 성능 대비 비용이 좋다는 평이 많아서, 클로드 소네트 같은 프리미엄 모델과의 격차를 어디까지 좁힐지 궁금합니다
287
Load more