Kilo Code
@kilocode.ai
Kilo is the fastest growing coding agent because it is open. #1 on OpenRouter. 500k+ Kilo Coders. 6.1 trillion tokens used per month.
Just a few days after the release of Gemini 3 - we're seeing that tool error rates are indeed significantly below what we see with Gemini 2.5.
But the king of tool use - Sonnet 4.5 - still has Gemini beat!
But the king of tool use - Sonnet 4.5 - still has Gemini beat!
November 20, 2025 at 6:50 PM
Just a few days after the release of Gemini 3 - we're seeing that tool error rates are indeed significantly below what we see with Gemini 2.5.
But the king of tool use - Sonnet 4.5 - still has Gemini beat!
But the king of tool use - Sonnet 4.5 - still has Gemini beat!
Did you know: Google's new IDE, Antigravity is a fork of Windsurf.
This makes it the most expensive proprietary fork in the history of tech.
Learn more:
blog.kilocode.ai/p/antigravit...
This makes it the most expensive proprietary fork in the history of tech.
Learn more:
blog.kilocode.ai/p/antigravit...
Antigravity is the most expensive PORK to date
PORK = a codebase forked from closed-source software
open.substack.com
November 20, 2025 at 6:16 PM
Did you know: Google's new IDE, Antigravity is a fork of Windsurf.
This makes it the most expensive proprietary fork in the history of tech.
Learn more:
blog.kilocode.ai/p/antigravit...
This makes it the most expensive proprietary fork in the history of tech.
Learn more:
blog.kilocode.ai/p/antigravit...
We used Gemini 3 Pro Preview inside Kilo Code to build a classic platform game with one prompt.
Then we shipped it in seconds with Kilo Deploy, and demoed it live during our webinar.
Try "Kilo Man" here:
kilomangame-uf1b1i.d.kiloapps.ai
Join our Discord to test Kilo Deploy:
discord.gg/pzzaJEeV
Then we shipped it in seconds with Kilo Deploy, and demoed it live during our webinar.
Try "Kilo Man" here:
kilomangame-uf1b1i.d.kiloapps.ai
Join our Discord to test Kilo Deploy:
discord.gg/pzzaJEeV
November 20, 2025 at 5:44 PM
We used Gemini 3 Pro Preview inside Kilo Code to build a classic platform game with one prompt.
Then we shipped it in seconds with Kilo Deploy, and demoed it live during our webinar.
Try "Kilo Man" here:
kilomangame-uf1b1i.d.kiloapps.ai
Join our Discord to test Kilo Deploy:
discord.gg/pzzaJEeV
Then we shipped it in seconds with Kilo Deploy, and demoed it live during our webinar.
Try "Kilo Man" here:
kilomangame-uf1b1i.d.kiloapps.ai
Join our Discord to test Kilo Deploy:
discord.gg/pzzaJEeV
Join us for a live Q&A on using AI coding tools in team environments.
Bring your questions about making AI coding work for your team—we'll tackle them live.
Bring your questions about making AI coding work for your team—we'll tackle them live.
November 17, 2025 at 7:38 PM
Join us for a live Q&A on using AI coding tools in team environments.
Bring your questions about making AI coding work for your team—we'll tackle them live.
Bring your questions about making AI coding work for your team—we'll tackle them live.
OpenAI just launched 4 new models: GPT-5.1, GPT-5.1 Chat, GPT-5.1-Codex and GPT-5.1-Codex-Mini.
Kilo Code Now Supports the Full GPT-5.1 Model Family.
Learn more:
blog.kilocode.ai/p/kilo-code-...
Kilo Code Now Supports the Full GPT-5.1 Model Family.
Learn more:
blog.kilocode.ai/p/kilo-code-...
Kilo Code Now Supports the Full GPT-5.1 Model Family
Here's how to get started.
blog.kilocode.ai
November 13, 2025 at 7:16 PM
OpenAI just launched 4 new models: GPT-5.1, GPT-5.1 Chat, GPT-5.1-Codex and GPT-5.1-Codex-Mini.
Kilo Code Now Supports the Full GPT-5.1 Model Family.
Learn more:
blog.kilocode.ai/p/kilo-code-...
Kilo Code Now Supports the Full GPT-5.1 Model Family.
Learn more:
blog.kilocode.ai/p/kilo-code-...
54% of devs resist AI due to control issues. The insight? AI is your new junior engineer - eager, fast, knowledgeable, and won't admit when it's wrong. Manage it like one: clear instructions, code reviews, gradual trust. Your senior skills just became more valuable
blog.kilocode.ai/p/how-to-bui...
blog.kilocode.ai/p/how-to-bui...
How To Build an AI Dev Team That Thinks, Builds, and Debugs Alongside Humans
AI adoption outside of tech: Enthusiastic.
blog.kilocode.ai
November 13, 2025 at 5:44 PM
54% of devs resist AI due to control issues. The insight? AI is your new junior engineer - eager, fast, knowledgeable, and won't admit when it's wrong. Manage it like one: clear instructions, code reviews, gradual trust. Your senior skills just became more valuable
blog.kilocode.ai/p/how-to-bui...
blog.kilocode.ai/p/how-to-bui...
We've reached a point where token-based pricing is CHEAPER than pseudo-credit based pricing.
To test our theory, we tested Augment Code's new credit system with 4 real coding tasks.
The results will surprise you: blog.kilocode.ai/p/testing-au...
To test our theory, we tested Augment Code's new credit system with 4 real coding tasks.
The results will surprise you: blog.kilocode.ai/p/testing-au...
Testing Augment Code's New Credit System with 4 Real Tasks
Have we reached a point where token-based pricing is becoming more economical than credit-based pricing?
blog.kilocode.ai
November 10, 2025 at 4:34 PM
We've reached a point where token-based pricing is CHEAPER than pseudo-credit based pricing.
To test our theory, we tested Augment Code's new credit system with 4 real coding tasks.
The results will surprise you: blog.kilocode.ai/p/testing-au...
To test our theory, we tested Augment Code's new credit system with 4 real coding tasks.
The results will surprise you: blog.kilocode.ai/p/testing-au...
Good news — MiniMax-M2 is back and free again on Kilo Code!
November 9, 2025 at 12:20 PM
Good news — MiniMax-M2 is back and free again on Kilo Code!
Still debugging Cursor’s pricing?
Get 2 months of Kilo Plans features for FREE when your team switches from Cursor to Kilo Code.
Learn more here: blog.kilocode.ai/p/kilo-for-c...
Get 2 months of Kilo Plans features for FREE when your team switches from Cursor to Kilo Code.
Learn more here: blog.kilocode.ai/p/kilo-for-c...
November 7, 2025 at 6:13 PM
Still debugging Cursor’s pricing?
Get 2 months of Kilo Plans features for FREE when your team switches from Cursor to Kilo Code.
Learn more here: blog.kilocode.ai/p/kilo-for-c...
Get 2 months of Kilo Plans features for FREE when your team switches from Cursor to Kilo Code.
Learn more here: blog.kilocode.ai/p/kilo-for-c...
58% of devs use AI tools that train on your code. The problem isn't AI adoption—your team's already doing that. The problem is shadow IT, scattered costs, and seniors becoming human linters for AI chaos. Time to move from indie subscriptions...
blog.kilocode.ai/p/your-team-...
blog.kilocode.ai/p/your-team-...
Your Team Is Already Using AI. Now What?
The AI coding revolution isn’t coming; it’s already in your codebase.
blog.kilocode.ai
November 7, 2025 at 2:55 PM
58% of devs use AI tools that train on your code. The problem isn't AI adoption—your team's already doing that. The problem is shadow IT, scattered costs, and seniors becoming human linters for AI chaos. Time to move from indie subscriptions...
blog.kilocode.ai/p/your-team-...
blog.kilocode.ai/p/your-team-...
Polaris Alpha just dropped—a stealth model that's completely free ($0 input/$0 output). 256K context, strong at coding + tool calling. The catch?
All prompts logged for model improvement. Free frontier-class inference with trade-offs.
All prompts logged for model improvement. Free frontier-class inference with trade-offs.
November 6, 2025 at 8:12 PM
Polaris Alpha just dropped—a stealth model that's completely free ($0 input/$0 output). 256K context, strong at coding + tool calling. The catch?
All prompts logged for model improvement. Free frontier-class inference with trade-offs.
All prompts logged for model improvement. Free frontier-class inference with trade-offs.
Earlier today, a small # of users experienced temporary issues accessing xAI models due to an upstream routing provider outage.
We quickly identified the cause and deployed an update to route requests directly to xAI. Most users saw little to no disruption.
Everything is now operating normally.
We quickly identified the cause and deployed an update to route requests directly to xAI. Most users saw little to no disruption.
Everything is now operating normally.
November 6, 2025 at 1:03 PM
Earlier today, a small # of users experienced temporary issues accessing xAI models due to an upstream routing provider outage.
We quickly identified the cause and deployed an update to route requests directly to xAI. Most users saw little to no disruption.
Everything is now operating normally.
We quickly identified the cause and deployed an update to route requests directly to xAI. Most users saw little to no disruption.
Everything is now operating normally.
Kilo Code’s new Parallel Agent mode lets you run multiple agents side by side from the CLI, all working together in real time.
No conflict. No switching. Just pure, parallel speed.
youtu.be/WYjxN8QUJi8
No conflict. No switching. Just pure, parallel speed.
youtu.be/WYjxN8QUJi8
Parellel Agents in Kilo Code
With Kilo Code’s Parallel Agents, you can now run multiple agents from the CLI. They can each handle their own task without overwriting or conflicting with each other.
Build, test, and debug in…
youtu.be
November 4, 2025 at 2:25 AM
Kilo Code’s new Parallel Agent mode lets you run multiple agents side by side from the CLI, all working together in real time.
No conflict. No switching. Just pure, parallel speed.
youtu.be/WYjxN8QUJi8
No conflict. No switching. Just pure, parallel speed.
youtu.be/WYjxN8QUJi8
One of the fastest growing open source models ever in Kilo Code - Minimax M2 took off in its first week of availability in the Kilo Code Provider.
Will be interesting to see what this week brings. Follow along here: kilocode.ai/models/minim...
Will be interesting to see what this week brings. Follow along here: kilocode.ai/models/minim...
November 3, 2025 at 6:22 PM
One of the fastest growing open source models ever in Kilo Code - Minimax M2 took off in its first week of availability in the Kilo Code Provider.
Will be interesting to see what this week brings. Follow along here: kilocode.ai/models/minim...
Will be interesting to see what this week brings. Follow along here: kilocode.ai/models/minim...
If you're part of an engineering team considering the switch to Kilo, join us next week for a live webinar on making the switch.
Open models
Open pricing
Open source
app.livestorm.co/kilocode/mak...
Open models
Open pricing
Open source
app.livestorm.co/kilocode/mak...
October 30, 2025 at 5:50 PM
If you're part of an engineering team considering the switch to Kilo, join us next week for a live webinar on making the switch.
Open models
Open pricing
Open source
app.livestorm.co/kilocode/mak...
Open models
Open pricing
Open source
app.livestorm.co/kilocode/mak...
Usage of MiniMax M2 is skyrocketing in Kilo Code!
Looking at early results, it seems error rates are stable with the other open source models on the market. This can often be a sticking point, so great to see the hard work of the MiniMax team paying off.
Looking at early results, it seems error rates are stable with the other open source models on the market. This can often be a sticking point, so great to see the hard work of the MiniMax team paying off.
October 30, 2025 at 4:44 PM
Usage of MiniMax M2 is skyrocketing in Kilo Code!
Looking at early results, it seems error rates are stable with the other open source models on the market. This can often be a sticking point, so great to see the hard work of the MiniMax team paying off.
Looking at early results, it seems error rates are stable with the other open source models on the market. This can often be a sticking point, so great to see the hard work of the MiniMax team paying off.
MiniMax-M2 now in Kilo Code: 10B active params (230B total) crushing SWE-Bench with compile-run-fix loops. Same intelligence as 70B+ models but 7x faster inference. Just tested: fixed a 3-file React bug in one pass.
What do you think? Check out more stats here: kilocode.ai/models/minim...
What do you think? Check out more stats here: kilocode.ai/models/minim...
October 29, 2025 at 5:12 PM
MiniMax-M2 now in Kilo Code: 10B active params (230B total) crushing SWE-Bench with compile-run-fix loops. Same intelligence as 70B+ models but 7x faster inference. Just tested: fixed a 3-file React bug in one pass.
What do you think? Check out more stats here: kilocode.ai/models/minim...
What do you think? Check out more stats here: kilocode.ai/models/minim...
No secret models
No "premium requests"
No rate limits
Complete control
Your productivity tool shouldn't be your biggest productivity problem. Join the Cursor Refugee program today. kilo.love/cursor
No "premium requests"
No rate limits
Complete control
Your productivity tool shouldn't be your biggest productivity problem. Join the Cursor Refugee program today. kilo.love/cursor
October 29, 2025 at 4:51 PM
No secret models
No "premium requests"
No rate limits
Complete control
Your productivity tool shouldn't be your biggest productivity problem. Join the Cursor Refugee program today. kilo.love/cursor
No "premium requests"
No rate limits
Complete control
Your productivity tool shouldn't be your biggest productivity problem. Join the Cursor Refugee program today. kilo.love/cursor
Now proud to present: Even more ways to connect to Kilo Code.
We now support Google, GitHub, GitLab, and LinkedIn sign-in — as well as SSO for our enterprise friends.
There's never been a better time to be a Kilo Coder!
We now support Google, GitHub, GitLab, and LinkedIn sign-in — as well as SSO for our enterprise friends.
There's never been a better time to be a Kilo Coder!
October 28, 2025 at 1:25 AM
Now proud to present: Even more ways to connect to Kilo Code.
We now support Google, GitHub, GitLab, and LinkedIn sign-in — as well as SSO for our enterprise friends.
There's never been a better time to be a Kilo Coder!
We now support Google, GitHub, GitLab, and LinkedIn sign-in — as well as SSO for our enterprise friends.
There's never been a better time to be a Kilo Coder!
Code Supernova is shutting down.
We found 2 models that are as good, if not even better:
1. Grok Code Fast 1 (still free)
2. Combine:
- GPT-5 Mini (for planning)
- Grok Code Fast 1 (for execution)
Read more: blog.kilocode.ai/p/code-super...
We found 2 models that are as good, if not even better:
1. Grok Code Fast 1 (still free)
2. Combine:
- GPT-5 Mini (for planning)
- Grok Code Fast 1 (for execution)
Read more: blog.kilocode.ai/p/code-super...
October 27, 2025 at 6:03 PM
Code Supernova is shutting down.
We found 2 models that are as good, if not even better:
1. Grok Code Fast 1 (still free)
2. Combine:
- GPT-5 Mini (for planning)
- Grok Code Fast 1 (for execution)
Read more: blog.kilocode.ai/p/code-super...
We found 2 models that are as good, if not even better:
1. Grok Code Fast 1 (still free)
2. Combine:
- GPT-5 Mini (for planning)
- Grok Code Fast 1 (for execution)
Read more: blog.kilocode.ai/p/code-super...
Webinar: Scaling AI Coding from Experiment to Integrated into Engineering Teams Workflow
Moving from "a few developers trying AI" to "AI as a standard part of how we build" is harder than it looks.
Moving from "a few developers trying AI" to "AI as a standard part of how we build" is harder than it looks.
October 27, 2025 at 5:39 PM
Webinar: Scaling AI Coding from Experiment to Integrated into Engineering Teams Workflow
Moving from "a few developers trying AI" to "AI as a standard part of how we build" is harder than it looks.
Moving from "a few developers trying AI" to "AI as a standard part of how we build" is harder than it looks.
The real cost of adding features? Keeping docs in sync.
Watch @olearycrew add a /config command:
- Kilo Code in VS Code: writes the implementation
- Kilo Code CLI: updates docs to match the new API
- Both running in parallel, sharing context
This is what AI orchestration actually looks like
Watch @olearycrew add a /config command:
- Kilo Code in VS Code: writes the implementation
- Kilo Code CLI: updates docs to match the new API
- Both running in parallel, sharing context
This is what AI orchestration actually looks like
October 23, 2025 at 5:31 PM
The real cost of adding features? Keeping docs in sync.
Watch @olearycrew add a /config command:
- Kilo Code in VS Code: writes the implementation
- Kilo Code CLI: updates docs to match the new API
- Both running in parallel, sharing context
This is what AI orchestration actually looks like
Watch @olearycrew add a /config command:
- Kilo Code in VS Code: writes the implementation
- Kilo Code CLI: updates docs to match the new API
- Both running in parallel, sharing context
This is what AI orchestration actually looks like
Kilo Code’s open-source AI coding agent is now available in your terminal!
→ Access hundreds of models
→ Switch between modes like ask, debug, architect, and orchestrate
→ Run autonomous mode to integrate with CI/CD environments
kilocode.ai/cli
npm install -g @kilocode/cli
→ Access hundreds of models
→ Switch between modes like ask, debug, architect, and orchestrate
→ Run autonomous mode to integrate with CI/CD environments
kilocode.ai/cli
npm install -g @kilocode/cli
October 22, 2025 at 3:27 PM
Kilo Code’s open-source AI coding agent is now available in your terminal!
→ Access hundreds of models
→ Switch between modes like ask, debug, architect, and orchestrate
→ Run autonomous mode to integrate with CI/CD environments
kilocode.ai/cli
npm install -g @kilocode/cli
→ Access hundreds of models
→ Switch between modes like ask, debug, architect, and orchestrate
→ Run autonomous mode to integrate with CI/CD environments
kilocode.ai/cli
npm install -g @kilocode/cli
We compared Claude Haiku 4.5 vs. GLM-4.6 vs. GPT-5 Mini.
Which one works the best for AI coding tasks?
Find the results here: blog.kilocode.ai/p/mini-model...
Which one works the best for AI coding tasks?
Find the results here: blog.kilocode.ai/p/mini-model...
Mini Models Battle: Claude Haiku 4.5 vs GLM-4.6 vs GPT-5 Mini
Which one works the best for coding tasks?
blog.kilocode.ai
October 21, 2025 at 9:12 PM
We compared Claude Haiku 4.5 vs. GLM-4.6 vs. GPT-5 Mini.
Which one works the best for AI coding tasks?
Find the results here: blog.kilocode.ai/p/mini-model...
Which one works the best for AI coding tasks?
Find the results here: blog.kilocode.ai/p/mini-model...