@kilocode @littledatauk. @stanford and @edinburghuni. Tech optimist. Former journalist.
- Tweets183
- Following219
- Followers125
- Likes398
ALT A simple partnership graphic on a light gray background. The Inception logo and wordmark appear at the top in black, with a large black “x” centered beneath it. Below, the Kilo Code logo is displayed inside a black rectangular banner with a bright yellow border and yellow text, indicating a collaboration between Inception and Kilo Code.
ALT Kilo's weekly New in Kilo graphic, week of June 15, 2026. Title reads "NEW IN KILO" with a strikethrough, replaced by "A WEEK OF SURPRISES" beside it. Below: ":: head spinning ::". Three items listed. 01: Fable 5, withdrawn by Anthropic, M3 and GLM 5.2 stepped in. 02: Grok Build 0.1, beat frontier models in Kilo Code Reviews. 03: Le Chaton Fat, crushed the benchmarks, hiding under the couch. Kilo logo top right, kilo.ai bottom left, @kilocode bottom right.
ALT Western illustration titled "On the Frontier: Building the Future, One Token at a Time." Four cowboys on horseback throw lassos at a galloping black horse labeled "Fable 5," which runs toward a wooden sign reading "Fable 5: Next-Gen Agentic Intelligence." The riders are labeled Kimi K2.7 Code ("Fast. Smart. Built to code."), GPT-5.5 ("Strong reasoning. Sharp every time."), Nemotron 3 Ultra ("Powerful. Precise. Unstoppable."), and MiniMax M3 ("Lightning fast. Built for scale."), set against a desert mesa landscape.
ALT Black-and-yellow comparison graphic from Kilo Code’s Code Audit Benchmark. The headline reads “MiniMax M3 vs Claude Opus 4.8.” Below, benchmark results show 17 bugs planted, with both models catching 13 bugs. Cost per run is highlighted as $0.07 for MiniMax M3 and $1.30 for Claude Opus 4.8. A note at the bottom reads, “Same codebase. Same prompt. Full breakdown in thread.” The Kilo Code logo appears in the upper-right corner.
ALT Graphic with black background and neon yellow text announcing “NEW IN KILO” for the week of May 18, 2026. The graphic highlights three updates in a ranked list format: “01 Grok Build 0.1” with the description “Makes it easy to build anything, fast.”, “02 Gemini 3.5 Flash” with the description “Leading a new type of flash model.”, and “03 Qwen 3.7 Max” with the description “Emerged much sooner than anyone expected.” On the right side, a highlighted “LEADER” section shows “Stepfun” with “30% of usage” and the note “Holding the top spot on Kilo Gateway this week.” Bottom left includes “kilo.ai | @kilocode” and bottom right features a small pixel-style logo icon.
ALT Benchmark comparison bar chart showing Kimi K2.6 (blue bars) against GPT-5.4 xhigh, Claude Opus 4.6 max effort, and Gemini 3.1 Pro thinking high (gray bars) across ten evaluations grouped into three categories. Kimi K2.6 leads on Humanity's Last Exam with tools (54.0 vs 52.1, 53.0, 51.4), BrowseComp is led by Gemini at 85.9 with Kimi at 83.2, DeepSearchQA f1-score led by Kimi at 92.5, Toolathlon led by GPT-5.4 at 54.6 with Kimi at 50.0, and OSWorld-Verified led by GPT-5.4 at 75.0 with Kimi at 73.1 under the General Agents category. Under Coding, Terminal-Bench 2.0 Terminus-2 is led by Gemini at 68.5 with Kimi at 66.7, SWE-Bench Pro led by Kimi at 58.6, and SWE-bench Multilingual led by Claude Opus at 77.8 with Kimi at 76.7. Under Visual Agents, MathVision with python is led by GPT-5.4 at 96.1 with Kimi at 93.2, and V* with python is led by GPT-5.4 at 98.4 with Kimi at 96.9. All scores are percentages.
ALT Kilo Code announcement graphic on black background with yellow accents. Yellow badge reads "Free for a limited time in Kilo." Headline: "Dola Seed 2.0 Pro — The Multimodal Leap." Five bullet points: matches GPT-5.2 and Gemini 3 Pro on benchmarks, true multimodal (images, video, visual documents), 256K context / 128K output, optimized for agentic workflows, native browser and computer use. Bottom bar reads "Available now in VS Code, CLI, KiloClaw" with kilo.ai branding.