FastGPT adds orchestration features

FastGPT said on September 7 that it had released v4.16.0 along with patch updates v4.8.10 and v4.9.2, adding conditional workflows, sandbox configuration changes and migration guides for self-hosted users. The company described the release as an update to its open-source AI orchestration platform for teams building workflow-based applications. (prnewswire.com) GitHub’s release page shows v4.16.2 as the latest tag as of September 3, following v4.16.1 and v4.16.0. FastGPT’s main repository had about 29,600 stars and 7,300 forks when checked on Tuesday. (github.com) ### What exactly changed in this release? FastGPT’s September 7 announcement said v4.16.0 introduced conditional workflows, which let teams route execution paths based on conditions inside a workflow. (fastgpt.io) The release also added sandbox configuration updates and migration guides aimed at helping users move older deployments onto the new version. FastGPT’s documentation says the 4.16 line also changed how agent sandboxes are managed. The company said the release refactored the sandbox lifecycle and migration flow with concurrency protection, resumable execution and idempotent retries for operations including creation, suspension, archiving, restoration and deletion. (prnewswire.com) ### Why are sandboxes getting so much attention? FastGPT’s technical documentation says OpenSandbox now keeps persistent volumes after stopping, allowing them to be reused on later runs. The company also said suspension and archive thresholds can now be controlled through environment variables. (prnewswire.com) The same documentation says App Chat will automatically disable sandbox functions if the agent sandbox is unavailable or unsupported by a team’s plan, while keeping models, tools, datasets and workflow nodes running. FastGPT said Apps and Skills now share runtime image upgrade status, and the Skill editor polls continuously for upgrade results instead of requiring manual refreshes. (fastgpt.io) ### What do the migration guides actually do? FastGPT’s 4.16.0 upgrade notes say the migration process first runs earlier normalization steps from v4.15.0-beta6. The documentation says that process fills in sourceType and sourceId for legacy sandboxes, removes obsolete fields, deletes orphaned resources that cannot be associated, and cleans up older Skill Debug Chat collections and bucket prefixes when sourceType is missing. (fastgpt.io) FastGPT’s technical center said the release also improved compatibility handling for legacy workflow data and tool parameters. The company said those changes were intended to reduce migration friction for older deployments. (fastgpt.io) ### How does this fit into FastGPT’s broader product push? FastGPT said in a separate release about two weeks earlier that v4.16.0 also brought user-level code sandboxes and knowledge-base metadata changes. The GitHub repository describes FastGPT as a platform for data processing, retrieval-augmented generation and visual AI workflow orchestration. (doc.fastgpt.cn) GitHub’s releases page shows FastGPT has continued shipping follow-on updates after the initial 4.16.0 release, with v4.16.1 and v4.16.2 posted afterward. That sequence suggests the company is still patching the branch as users adopt the new workflow and sandbox behavior. (fastgpt.io) That is an inference based on the release chronology. ### Where can users find the next deployment steps? FastGPT’s technical center links three follow-up documents for 4.16.0 users: a guide to migrate agent sandbox data, a guide to configure the agent sandbox proxy, and bug-fix documentation for the 4.16.0 release. The GitHub release page shows v4.16.2 as the latest published version, and FastGPT’s documentation site hosts the upgrade instructions for the 4.16 branch. (prnewswire.com) (fastgpt.io) (github.com)