English translation is being prepared.
Matched fieldsSummary
←Search results
Search “Open-source ecosystem”
English translation is being prepared.
Matched fieldsSummary
Modal releases open-source training server Spindle, compatible with the Tinker API
Modal releases the open-source training server Spindle, which can run in your own Modal workspace, is compatible with the Tinker API, and bills by GPU time rather than tokens. Deployment only requires running spindle deploy, and it supports full-parameter training and multiple LoRAs, covering a variety of open-source models. In the official comparison, a single-round DAPO-math task with 8-card H200 training, 8-card H200 inference, and 8 clients sharing costs $1.99 per million output tokens, 80% lower than Tinker.
Modal releases the open-source training server Spindle, which can run in your own Modal workspace, is compatible with the Tinker API, and bills by GPU time rather than tokens. Deployment only requires running spindle deploy, and it supports full-parameter training and multiple LoRAs, covering a variety of open-source models. In the official comparison, a single-round DAPO-math task with 8-card H200 training, 8-card H200 inference, and 8 clients sharing costs $1.99 per million output tokens, 80%
Matched fieldsBody
Alibaba Tongyi MAI-UI team open-sources general computer-use agent tech stack OpenGCUA
Alibaba Tongyi MAI-UI team open-sources the complete general computer-use agent tech stack OpenGCUA, including harness, cross-device SDK, native macOS app, evaluation trajectories, and reference agent implementation. Its core design is that code execution and decision-making shift from the harness to the model; the model takes screenshots on demand, writes Python programs with loops and conditional judgments to orchestrate GUI operations, API calls, and shell commands, and maintains its own todo list. In evaluations of 8 frontier models on OSWorld-v2, the average partial score rose from 45.3 to 64.9, and the average number of steps per task decreased by 53%; in real-machine macOS tests, under the same harness, Qwen3.8-Max's inference token consumption was 19.4 times that of GPT-6 Astra, completing 3 of 4 tasks.
Alibaba Tongyi MAI-UI team open-sources the complete general computer-use agent tech stack OpenGCUA, including harness, cross-device SDK, native macOS app, evaluation trajectories, and reference agent implementation. Its core design is that code execution and decision-making shift from the harness to the model; the model takes screenshots on demand, writes Python programs with loops and conditional judgments to orchestrate GUI operations, API calls, and shell commands, and maintains its own todo
Matched fieldsBody
Cloudflare Open-Sources Void to Help IT Manage Employee Vibe-Coding Apps
Cloudflare announced the open-sourcing of Void, which includes three components: Void SDK, Void CLI, and Void Platform. Enterprises can self-host the platform to deploy and manage apps built by employees with AI, and control who can deploy and access them. Void 1.0 is released together with the complete source code, and the SDK and CLI were previously published via npm under the MIT license. Self-hosting requires the paid Cloudflare Workers for Platforms plan, starting at $25 per month, plus R2 storage and a separately hosted PostgreSQL database; there is currently no fully managed version.
Cloudflare announced the open-sourcing of Void, which includes three components: Void SDK, Void CLI, and Void Platform. Enterprises can self-host the platform to deploy and manage apps built by employees with AI, and control who can deploy and access them. Void 1.0 is released together with the complete source code, and the SDK and CLI were previously published via npm under the MIT license. Self-hosting requires the paid Cloudflare Workers for Platforms plan, starting at $25 per month, plus R2
Matched fieldsBody