DataRobot
DataRobot is rebuilding itself as the governance and capacity layer under everyone else's agents
A side-by-side editorial comparison of Docling and LiveKit Agents — release velocity, themes, recent moves, and the top alternatives to consider.
Docling keeps swallowing new formats, and now the parsing engines behind them are swappable.
Docling releases every three to four days, alternating feature drops with tight fix releases. The current one is purely corrective: DOCX headings detected by outline level when the style is not literally named Heading, Markdown tables keeping their last cell without a trailing pipe, and the service client serializing engine options in full. Format coverage now spans PDF, Office, ODF, HTML, JATS, email, audio and video.
LiveKit Agents ships weekly, and the work is nearly all provider breadth and failure paths.
The two newest entries are bare release tags — 1.6.10 and a 1.6.11 release candidate — with no notes. The releases with content follow one pattern: add speech and avatar vendors (Deepgram Flux, Bland TTS, Spatius avatars, Phonic and Smallest STT options), then fix the ways conversations go wrong. Endpointing and turn detection recur constantly — max_delay held constant under dynamic endpointing, primary speaker preserved through unattributed segments, an Ink-2 speech-onset delay that pushed back STT interruptions. 1.6.8 deprecates the console and dev modes in favour of the lk agent CLI.
Docling releases every three to four days, alternating feature drops with tight fix releases. The current one is purely corrective: DOCX headings detected by outline level when the style is not literally named Heading, Markdown tables keeping their last cell without a trailing pipe, and the service client serializing engine options in full. Format coverage now spans PDF, Office, ODF, HTML, JATS, email, audio and video.
The engine layer is where the interesting movement is. Docling is shifting from one opinionated pipeline to a set of interchangeable layout, table and OCR backends the caller picks per run, which turns the library into a harness for models rather than a fixed parser. A second thread: the project shipped agent skills for itself in v2.118.0 and a separate docling-client package in v2.120.0, both pointing at being consumed programmatically rather than only imported. The structural-inference work — heading levels from font weight, now from DOCX outline levels — shows the parser learning to read documents that never declared their own structure.
Expect the engine-selection surface to keep widening, with OCR joining layout and table structure as a CLI-selectable backend. The steady stream of format-specific crash fixes suggests coverage is outrunning hardening, so more of these short corrective releases are likely between feature drops.
The two newest entries are bare release tags — 1.6.10 and a 1.6.11 release candidate — with no notes. The releases with content follow one pattern: add speech and avatar vendors (Deepgram Flux, Bland TTS, Spatius avatars, Phonic and Smallest STT options), then fix the ways conversations go wrong. Endpointing and turn detection recur constantly — max_delay held constant under dynamic endpointing, primary speaker preserved through unattributed segments, an Ink-2 speech-onset delay that pushed back STT interruptions. 1.6.8 deprecates the console and dev modes in favour of the lk agent CLI.
This is a maintenance train rather than a roadmap: breadth across providers plus relentless correctness work on turn-taking, which is where voice agents actually fail. The one demand on users in the window is the CLI migration. Cadence is roughly weekly and the contributor list is largely external — vendors landing their own plugins.
More provider plugins and more endpointing fixes; the deprecated console and dev modes are the next thing likely to be removed outright.
Other ai-assistants 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 Docling or LiveKit Agents.
DataRobot is rebuilding itself as the governance and capacity layer under everyone else's agents
Snorkel has stopped labeling data and started defining what agent competence means.
NEURONwriter is publishing the AI-search playbook faster than it is shipping the tool.
D-ID's feed is comparison marketing, with simpleshow folded into the pitch
Pictory publishes usage data from 1.5 million videos, but its feed carries no releases
OpenRouter's feed turns to documentation of the routing and image work it already shipped
See all Docling alternatives → · See all LiveKit Agents alternatives →
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
They serve adjacent needs but don't currently overlap on shipped themes. Docling is currently shipping more aggressively (velocity 6.3 vs 5.0), with 0 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. Docling is currently shipping more aggressively (velocity 6.3 vs 5.0), with 0 editorial sparks in the last 30 days against 0. For your specific use case, the alternatives sections above list other ai-assistants products to evaluate alongside.
Top Docling alternatives in ai-assistants are ranked by recent ship velocity. Browse the "Docling alternatives" section above for the current picks, or visit /alternatives/docling for the full list with editorial commentary on each.
Top LiveKit Agents alternatives in ai-assistants are ranked by recent ship velocity. Browse the "LiveKit Agents alternatives" section above for the current picks, or visit /alternatives/livekit-agents for the full list with editorial commentary on each.