ESPHome
ESPHome 2026.9.0 ships template climate component, OTA encryption with API key, and ESP-NOW for ESP32-P4
A side-by-side editorial comparison of Expo and Lima — release velocity, themes, recent moves, and the top alternatives to consider.
| Feature | Expo | Lima |
|---|---|---|
| Sector | Infra & APIs | Infra & APIs |
| Velocity score | 6.3 | 2.5 |
| Sparks · 30d | 1 | 0 |
| Top themes | sdk, mobile-devtools, observability, eas | virtual-machines, containers, windows-support, cross-platform |
| Last editorial update | 3h ago | 15d ago |
| Website | — | Visit → |
Expo kills its AI agent, SDK 58 beta arrives as EAS observability stack hits GA
Expo is at a turning point: the EAS infrastructure layer — build, update, and now observe — is maturing into production-grade tooling, with EAS Observe reaching GA in August and SDK 58 entering beta this week. Simultaneously, Expo shut down its AI Agent experiment after months in closed beta, a signal that building an AI coding assistant directly on the platform didn't reach product-market fit. The MCP server and Claude desktop integration indicate where the AI bet now sits: in connectors, not agents.
Lima keeps widening which machines it can run on, and run.
Lima runs Linux virtual machines for container workloads, and its 2.x line has been a steady expansion of the host and guest matrix. macOS and FreeBSD guests arrived at 2.1.0, Windows Server 2025 and Windows 11 guests followed through 2.2.0, and 2.3.0-beta.0 adds an experimental HCS driver for Windows hosts along with further Windows guest work. Each platform addition lands marked experimental first, and much of the surrounding work is CLI and template maintenance.
Expo is at a turning point: the EAS infrastructure layer — build, update, and now observe — is maturing into production-grade tooling, with EAS Observe reaching GA in August and SDK 58 entering beta this week. Simultaneously, Expo shut down its AI Agent experiment after months in closed beta, a signal that building an AI coding assistant directly on the platform didn't reach product-market fit. The MCP server and Claude desktop integration indicate where the AI bet now sits: in connectors, not agents.
Expo is consolidating around EAS as an observable infrastructure layer, not as an AI product builder. The MCP server and third-party integrations (PostHog, Claude) suggest the strategy is to serve as a data and build substrate that AI tools operate on top of. SDK 58 will push new device capabilities through the standard pipeline; EAS Observe data becoming a programmable surface for AI tools is the natural next extension of this approach.
The next move is a stable SDK 58 release and deeper AI-tool integrations via MCP — Expo is betting its AI story on being the substrate that coding assistants operate on, not on shipping its own. Expect more one-command observability integrations to follow PostHog.
Lima runs Linux virtual machines for container workloads, and its 2.x line has been a steady expansion of the host and guest matrix. macOS and FreeBSD guests arrived at 2.1.0, Windows Server 2025 and Windows 11 guests followed through 2.2.0, and 2.3.0-beta.0 adds an experimental HCS driver for Windows hosts along with further Windows guest work. Each platform addition lands marked experimental first, and much of the surrounding work is CLI and template maintenance.
The project is converging on a single tool that runs any guest on any host, and it is doing so from both ends: guest support and host drivers advance in alternating releases. Windows is the current focus on both sides, having moved from an unsupported host default to qemu and now to a native HCS driver option. The LFX mentee contributions behind much of this suggest the platform matrix expands as contributor capacity allows rather than on a roadmap.
The HCS driver should follow the same path the other platform additions took, moving from experimental to a supported or default option on Windows hosts once the guest work stabilises around it.
Other Infra & APIs products tracked by Sparkpulse, ranked by recent ship velocity. Each card links to a full editorial trajectory and lets you pivot into a head-to-head comparison with either Expo or Lima.
ESPHome 2026.9.0 ships template climate component, OTA encryption with API key, and ESP-NOW for ESP32-P4
Redocly ships a built-in MCP server page across its entire docs platform, with public and authenticated endpoints
Jackett in pure tracker-maintenance mode, daily domain and category updates only
Skipper adds Redis as a distributed L2 cache backend amid a high-frequency patch cadence.
Kubernetes v1.37 matures its memory management and scheduling stack for AI/ML workloads.
GitHub Copilot gets cost-aware inference tiers as enterprise AI tooling tightens across the platform.
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
They serve adjacent needs but don't currently overlap on shipped themes. Expo is currently shipping more aggressively (velocity 6.3 vs 2.5), with 1 editorial sparks in the last 30 days against 0. See the at-a-glance table above for a side-by-side breakdown of velocity, recent sparks, and editorial themes.
Sparkpulse doesn't pick a winner — we score release velocity, not feature parity. Expo is currently shipping more aggressively (velocity 6.3 vs 2.5), with 1 editorial sparks in the last 30 days against 0. For your specific use case, the alternatives sections above list other Infra & APIs products to evaluate alongside.
Top Expo alternatives in Infra & APIs are ranked by recent ship velocity. Browse the "Expo alternatives" section above for the current picks, or visit /alternatives/expo for the full list with editorial commentary on each.
Top Lima alternatives in Infra & APIs are ranked by recent ship velocity. Browse the "Lima alternatives" section above for the current picks, or visit /alternatives/lima for the full list with editorial commentary on each.