Developers do not adopt coding agents because a benchmark says a model is powerful. They move from autocomplete to chat, then to single-agent tasks and multi-agent orchestration when each step proves reliable, understandable and recoverable. Telemetry from Kilo Code’s large but early-adopter-heavy user base shows that trust, context quality, model choice and feedback loops—not raw scores alone—determine how much autonomy teams will allow.
What the token data actually covers
Brendan O’Leary, founding developer relations engineer at Kilo Code, presented the findings in a QCon AI session recorded June 2, 2026. InfoQ published a transcript on October 15, 2026. The session describes Kilo Code as an open-source agentic engineering tool available through a command-line interface, Visual Studio Code, JetBrains products and cloud-agent surfaces.
As an Amazon Associate I earn from qualifying purchases.
The conference abstract describes more than 25 trillion tokens across more than 1.5 million developers. By the time of the recording, O’Leary said the gateway had processed over 40 trillion tokens. Those figures are cumulative usage totals, not a count of unique production deployments or a measure of code quality.
The telemetry comes from Kilo Code’s gateway, which gives users access to more than 500 models. The analysis requested-model data, removed anonymous and internal users, filtered automated traffic and excluded autocomplete from most analyses. It is therefore not a random sample of the software industry: the cohort is disproportionately made up of early adopters who deliberately chose open-source, configurable tooling.
#1 Best Overall
- [Ultra-Lightweight Ear Buds Designed for Small Ears] Each earbud weighs only 3.7g and features a compact, ergonomic in-ear design made especially for small ears. Secure, low-profile, and comfortable for workouts, all-day wear or overnight sleeping.The ultra-slim profile sits flush in your ear with no hard edges poking or rubbing, so you can even sleep on your side without discomfort.
- [Immersive Stereo Sound with TOZO OrigX Technology] TOZO OrigX tuning delivers clear vocals, balanced mids, and natural stereo sound for music, podcasts, and videos.
- [Long Battery Life for Daily Use] Get up to 7 hours of playtime on a single charge, with up to 32 hours total using the charging case—ideal for workdays, commuting, and extended listening sessions.
- [Bluetooth 5.3 & Stable Connection] Bluetooth 5.3 provides fast pairing, stable wireless performance, and reduced dropouts as you move around home or office.
- [Deep Bass with Clear Vocals] High-performance drivers produce punchy bass while keeping vocals clean and detailed for everyday listening.
That makes the data valuable as real-world product telemetry, but it does not establish an industry-wide adoption rate, a causal link between a model and developer productivity, or reliable differences by programming language, department or task.
Adoption is a ladder, not a switch
O’Leary’s central model is a progression in which reliability at one rung determines whether developers will permit the next level of autonomy.
Rank #2
- 【IP68 Waterproof Multifunctinal Smart Watch】Torjalph fitness activity tracker smart watch is IP68 waterproof technology ,you can wear it when swimming,handwashing,it's rain-proof which means that you perform multiple exercise modes outside without worrying about sweat or rain corroding your fitness smart watch,and this watch equipped with a 230 mAh large capacity battery magnetic charging ,and this sports activity tracker watch can be used 5-7 days If fully charged.
- 【Larger 1.83'' Full Screen Watch & Multi-dial Selection】1.83 inch full HD color touch screen smart watch for men / women provides excellent 240*284 resolution image quality ,This Bluetooth smart watch offers a better touch experience and remarkable immersion and ,more than 200 watch faces for you to choose from App"Da Fit ", you can change the dial style according to your mood to wear keeping the world on your wrist
- 【Accurate Activity Tracker Smart Watch 】sports watch activity tracker uses sweat and dust resistant breathable silicone strap, our Fitness Tracker Smart Watch for Android iOS Phones supports 12 sports modes,walking,running,bicycle,swimming,badminton,basketball,football,rope skipping. This bluetooth calls smart watch for men women is lightweight and stylish, bringing more convenient and relaxing sports experience while also providing more protection.
- 【Your 24 Hour Personal Daily Health Monitor】Fitness watch for women men full touch screen IP67 waterproof bluetooth smartwatch for Android iOS Phones with heart rate blood pressure monitor multifunction Andriod smart watch with steps counter sleep monitor sports fitness watch, monitoring blood oxygen saturation allows you to understand your physical condition in real time to track fitness progress ,to know your health status in time.
- 【More Functions Bluetooth Call Smart Watch】This fitness activity smart watch full screen for iPhone and Android phones includes a bunch of new features including:The bluetooth smart watch can hang up the phone, receive and display WhatsAPP /Text /Facebook /Twitter /Messenger reminder.Support sedentary reminder, alarm clock, stopwatch, timer,weather information,music control, remote camera control compatible with Android 5.0 and above or iOS 9.0 and above smartphones.
| Stage | What the system does | Typical feedback loop | What must improve before advancing | Failure cost |
|---|---|---|---|---|
| Autocomplete | Suggests a small code continuation while a developer types | Immediate | Low-latency suggestions and easy rejection | Usually limited to a few keystrokes or a local edit |
| Chat | Answers questions or proposes code after a prompt | Seconds to minutes | Useful context and an answer the developer can inspect | Wasted attention or an incorrect snippet |
| Single agent | Plans and executes a bounded coding task | Minutes | Reliable retrieval, tool permissions and visible progress | Wrong files changed, tests missed or time lost debugging the agent |
| Agent orchestration | Coordinates multiple agents or longer-running subtasks | Longer and less continuous | Strong routing, checkpoints and recovery signals | Compounded errors with too little information to find the first wrong decision |
Autocomplete can feel dependable because the result is visible almost immediately. Orchestration removes the developer from more of the loop, so a team needs better evidence—not merely a more capable model—to know whether work is proceeding correctly.
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Why benchmark scores fail to predict adoption
Benchmarks measure performance under a defined test setup. Adoption is a workflow decision involving a repository’s conventions, the amount of context available, latency, cost, permissions and the team’s ability to inspect and undo changes. A model can score well and still be a poor fit if it receives incomplete files, takes too long, or produces decisions developers cannot evaluate.
Rank #3
- [Low Noise High Fidelity] New bee guitar cord with 22AWG oxygen-free-copper center conductors provides enhanced signal transfer, the PVC insulation material and the high-density stranded copper double shield provide noise rejection for quiet operation
- [Ideal for Choice] New bee 1/4 straight and right angle metal connectors, ensures a secure connection in any instrument, pedal, or amp jack. L shape great fro on stage guitar stands
- [Wide Compatible] Suitable for mixer, power amplifier, microphone, effector, decoder, equalizer, electoronic organ, electric guitar, electric bass and other electronic equipments. Widely used in the installation of smart home, karaoke speaker, professional stage audio, etc
- [Easy Disassemble] Making incidental repairs quick and easy! Very nice, flexible cable
- [Dont Hesite to Order] 24 hours customer services and professional technology team are standing by
The observed gateway behavior illustrates that model choice is becoming less concentrated. Over the eight weeks leading through April 2026, Anthropic’s share of Kilo Code gateway requests fell from 11.2% to 6.5% while both users and request volume grew threefold. Those percentages describe this specific gateway cohort and period; they are not a market-share estimate. The directional lesson is that developers increasingly mix providers and models rather than treating one model as a permanent default.
For an engineering organization, a useful evaluation asks whether a tool completes representative tasks with an acceptable review burden. Record the task’s starting context, the files retrieved, model and provider selected, time to useful feedback, test results, human edits and rollback rate. A benchmark score can be one input, but it cannot replace that workflow evidence.
Rank #4
- Digital Family Calendar: BSIMB digital calendar wall touch screen puts an end to your chaotic life. Designed for busy families, the digital calendar keeps you organised and in step with your family. Assign different colours and tasks to each family member to make family scheduling easier and more intuitive.
- Easy Setup and Auto-Sync: Plug in, connect to WLAN and auto-syn your electric calendar in minutes. Digital wall calendar compatible with Google calendar, iCloud calendar, Outlook calendar, Cozi calendar and Yahoo calendar. Add events directly on the device or through the ecalendar app. Electronic calendar stays current whether you're at work or traveling—no manual updates needed.
- All-In-One Home Assistant: Smart calendar builds habits and encourage independence with personalised chores. Display daily meal plan to reduce the daily ‘What's for dinner?’ stress. To-do lists keep track of everything you want. Turning chores into valuable achievements with stars and rewards, making life easier and more enjoyable.
- Manage Anytime Anywhere: BSIMB electronic calendar for family stay connected with the powerful ecalendar app. Access and manage your digital wall calendar anytime, anywhere with free features like stars & rewards, photo screensaver, chore charts, meals plans, to-do lists etc. Electronic planner paid features like Magic Import tailor advanced tools for users with special needs.
- Stay in Touch with Loved Ones: Share photos or videos from your phone to electronic wall calendar with app, instantly update your life moment, and keep in touch with your family, especially those who are far away; you can also surprise your smart calendar digital wall by sending him a wish card; it's also great for keeping tech-illiterate elders in touch with the youngsters.
The three technical breakpoints that decide trust
Context construction is infrastructure
An agent can only reason over the information it receives. Retrieval must consistently supply the relevant files, repository conventions, interfaces, tests and task requirements while excluding distracting or stale material. O’Leary calls a well-designed retrieval system the “number one thing to get right” and argues for “treating context as infrastructure, not as a feature.”
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Fix the driver behind crashes, sound loss and screen glitches3Clear out junk files and repair common Windows errors- Define what repository and task signals are required for each class of work.
- Make retrieved files and instructions inspectable before or during execution.
- Keep conventions, dependency boundaries and test commands current.
- Measure whether the agent used the right context, not just whether its final patch compiled.
Model routing should be provider-independent
Different tasks favor different models. A routing layer can select among more than 500 available models in Kilo Code’s gateway, rather than hard-wiring every task to one provider. Routing should account for task complexity, context length, latency, tool support and failure history. Keeping the interface provider-independent also makes it possible to change models without redesigning the entire workflow.
Feedback must expose the path, not only the patch
As autonomy increases, a final diff is too late and too thin a signal. Developers need intermediate progress: which context was selected, what plan was formed, which tools ran, what tests failed and why the agent changed direction. O’Leary summarizes the requirement as: “Your tooling has to surface what agents decided not just what they did.” Without that trail, a team may see an incorrect result but have no practical way to locate the first bad assumption.
How to roll out agentic coding without stalling at chat
- Choose bounded, reversible work. Start with tasks that have clear acceptance tests, limited write scope and an easy rollback path.
- Recruit patient early users. Begin with developers who are motivated to iterate through ambiguity and willing to document failures. Their successful examples provide social proof for more skeptical colleagues.
- Instrument every rung. Capture retrieval quality, model routing, latency, tool calls, test outcomes, human edits and rollback events. Do not rely on anecdotes alone.
- Add checkpoints before expanding permissions. Require review or approval before broad file changes, production-affecting actions or long-running orchestration.
- Publish working patterns. Senior engineers often benefit when they understand the system and can provide precise context. Reusable repository instructions, task templates and test conventions make agent work more predictable.
- Promote autonomy only after recovery is proven. Move from chat to single-agent execution, then to orchestration, when users can detect wrong turns early and recover without disproportionate effort.
Trust is asymmetric: many successful interactions can be undone by one visible failure. A staged rollout limits the blast radius while the organization improves retrieval, routing and observability.
What this evidence cannot answer
- It does not show what percentage of all developers use autonomous coding agents.
- It does not prove that one model or provider causes higher productivity.
- It does not provide robust comparisons across departments, languages or task categories.
- It offers only qualitative observations about senior engineers, not a controlled measurement of their performance.
- Because the cohort sought open-source configurable tooling, results may differ substantially in organizations using other procurement, security or platform arrangements.
The strongest conclusion is narrower and more actionable: adoption follows earned confidence. Teams that improve context construction, preserve model choice and expose intermediate decisions give developers the evidence needed to grant an agent more autonomy.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
