Remote-first AI coding startup Kilo doesn’t think software developers should pledge their unwavering allegiance to any one development environment — and certainly not to any one model or harness.
This week, the startup backed by GitLab co-founder Sid Sijbrandij unveiled Kilo CLI 1.0, a complete rebuild of its command-line tools that offers support for more than 500 different built-in AI models from proprietary leaders and open source rivals like Alibaba’s Quon.
This comes just weeks after Kilo launched Slackbot, which allows developers to ship code directly from Salesforce’s popular messaging service (Slack, which VentureBeat also uses), powered by Chinese AI startup Minimax.
This release marks a strategic pivot away from the IDE-centric “sidebar” model popularized by industry giants like Cursor and GitHub Copilot, or dedicated apps like the new OpenAI Codex, and even terminal-based rivals like Codex CLI and Cloud Code, which aim to embed AI capabilities into every piece of professional software workflows.
By launching a model-agnostic CLI along the lines of its Slack bot, Kilo is making a calculated bet: The future of AI development is not about a single interface, but about tools that travel with engineers between IDEs, terminals, remote servers, and team chat threads.
In a recent interview with VentureBeat, Kilo CEO and co-founder Scott Breitnoder explained the need for this fluidity: “This experience feels a little too fragmented right now… As an engineer, sometimes I’m going to use the CLI, sometimes I’m going to go into VS Code, and sometimes I’m going to pull an agent out of Slack, and not have people running around.”
He said Kilo CLI 1.0 is “built specifically for this world…for the developer who bounces between their local IDE, a remote server via SSH, and a terminal session at 2 a.m. to fix production bugs.”
Technology: Rebuilt for ‘Kilo Speed’
Kilo CLI 1.0 is a fundamental architectural change. While 2025 was the year senior engineers started taking AI vibe coding seriously, Kilo believes 2026 will be defined by the adoption of agents that can independently manage end-to-end tasks.
The new CLI is built on an MIT-licensed, open-source foundation, specifically designed to function in the terminal sessions where developers often find themselves during critical production events or intensive infrastructure work.
For Breitenother, building in the open is non-negotiable: “When you build in the open, you build better products. You get this big circle of contributors… Your community isn’t just passive users. They’re actually part of your team helping you grow your product… Honestly, some people might say open source is a weakness, but I think it’s our superpower.”
The core of this “agent” experience is Kilo’s ability to move beyond simple auto-completion. The CLI supports several operating modes:
- Code Mode: For high-speed generation and multi-file refactors.
-
Architect Mode: For high-level planning and technical strategy.
-
debug mode: For systematic problem diagnosis and solution.
solving multi-session memory
To solve the persistent problem of “AI amnesia” – where an agent loses context between sessions – Kilo uses a “memory bank” feature.
This system maintains state by storing references in structured Markdown files within the repository, ensuring that an agent working in the CLI has the same understanding of the codebase as an agent working in a VS Code sidebar or Slack thread.
The synergy between the new CLI and “Kilo for Slack” is central to the company’s “Agent Anywhere” strategy. Launched in January, the Slack integration allows teams to fix bugs and push pull requests directly from conversations.
Unlike competing integrations from Cursor or Cloud Code – Kilo claims to be limited by a single-repo configuration or lack of persistent thread state – Kilo’s bot can pull references from multiple repositories simultaneously.
“Engineering teams don’t make decisions in the IDE sidebar. They make them in Slack,” Breitnoder emphasized.
The ‘superpower’ of extensibility and open source
An important component of Kilo’s technical depth is its support for the Model Context Protocol (MCP). This open standard allows Kilo to communicate with external servers, expanding its capabilities beyond local file manipulation.
Through MCP, Kilo agents can integrate with custom tools and resources, such as internal documentation servers or third-party monitoring tools, effectively turning the agent into a specialized member of the engineering team.
This extensibility is part of KILO’s broader commitment to model agnosticism. While Minimax is the default for Slack, the CLI and extensions support a huge range of over 500 models, including Anthropic, OpenAI, and Google Gemini.
Pricing: The economics of ‘per dollar AI output’
Kilo is also attempting to disrupt the economics of AI development with “Kilo Pass,” a subscription service designed for transparency.
The company charges exact provider API rates with zero commission – $1 of Kilo Credit equals $1 of provider cost.
Breitenother is critical of the “black box” subscription models used by others in the sector: “We’re selling infrastructure here… You hit some kind of arbitrary, vague line, and then you start to be strangled. The world won’t work that way.”
The Kilo Pass tier offers “speed rewards” providing bonus credits for active customers:
- Starter ($19/month): Up to $26.60 in credits.
-
Pro ($49/month): Up to $68.60 in credits.
-
Expert ($199/month): Up to $278.60 in credits.
To encourage early adoption, Kilo is currently offering a “Double Welcome Bonus” until February 6, giving users 50% free credit for their first two months.
For power users like Sylvain, this flexibility is a major attraction: “Kilo Pass is exactly what I was waiting for. I can use my credits when I need them and save them when I don’t – it finally fits how I really use AI.”
Community, security and competition
The arrival of Kilo CLI 1.0 puts it in direct conversation with terminal-native heavyweights: Anthropic’s Cloud Code and Block’s Goose.
Outside of the terminal, in the more full-featured IDE space, OpenAI recently launched a new Codex desktop app for macOS.
Cloud Code offers a highly sophisticated experience, but it comes with vendor lock-in and high costs – up to $200 per month for tiers that still include token-based usage caps and rate limits. Independent analysis shows that these limits are often eliminated within minutes of intensive work on large codebases.
OpenAI’s new Codex app similarly supports a platform-locked approach, acting as a "Command Center for Agents" Which allows developers to monitor independently running AI systems for up to 30 minutes.
While codecs offer powerful features like "Skill" Connecting to tools like Figma and Linear, it is fundamentally designed to protect OpenAI’s ecosystem in a highly competitive market.
In contrast, Kilo CLI 1.0 uses an MIT-licensed OpenCode foundation to provide a production-ready terminal user interface (TUI) that allows engineers to swap between 500+ models.
This portability allows teams to select the best cost-to-performance ratio – perhaps using the lightweight model for documentation but swapping to the frontier model for complex debugging.
Regarding security, Kilo ensures that models are hosted on US-compliant infrastructure such as AWS Bedrock, allowing proprietary code to remain within the trusted perimeter while taking advantage of the most efficient intelligence available.
Goose offers an open-source alternative that runs completely for free on the user’s local machine, but feels more localized and experimental.
Kilo positions itself as a middle ground: a production-hardened tool that maintains open-source transparency while providing infrastructure at scale across an enterprise.
This contrasts with the broader market’s dual-use concerns; While OpenAI creates sandboxes to secure autonomous agents, Kilo’s open-core nature allows "super power" Community audit and level of contribution.
The future: a ‘make suit’ for the brain
With $8 million in seed funding and a “right of first refusal” agreement with GitLab that runs until August 2026, Kilo is positioning itself as the backbone of the next generation developer stack.
Breitenother sees these devices as “exoskeletons” or “mach suits” for the brain rather than replacements for human engineers.
“We’ve really made our engineers product owners,” Breitner revealed. “The moment they’re freed from writing code, they’re actually doing a lot of thinking. They’re setting the strategy for the product.”
By opening up the engineering stack – separating the agentic interface from the model and the model from the IDE – Kilo offers a roadmap to a future where developers think architecturally while machines build the structure.
Breitenother concluded, “I think it’s the closest thing to magic we can encounter in our lives.” For those wanting “kilo speed”, the IDE sidebar is just the beginning.
<a href