Tigris
An engineering blog making one argument: an agent's working state belongs in a forkable bucket
Comparison · DevOps
Refine vs Zed
A side-by-side editorial comparison of Refine and Zed — release velocity, themes, recent moves, and the top alternatives to consider.
Refine vs Zed: at a glance
What is Refine?
Refine's v5 majors are out; what's shipping now is the bug tail, mostly from contributors
Refine is a React framework for building internal tools and admin panels, and this window is entirely patch releases against the already-shipped v5 core and v8 MUI packages. The fixes cluster in three places: URL state parsing under syncWithLocation, access-control metadata not reaching button hooks, and CLI binary resolution inside monorepo workspaces.
What is Zed?
Zed ships weekly, and each week the agent panel and the Git surface get another turn of the crank.
Zed runs a weekly stable release with point releases in between that exist largely to pick up newly launched models — Claude Opus 5 and the GPT 5.6 family each got a same-week patch. The substantive work splits in two: a Git surface being built toward complete in-editor staging and review (hunk-level multibuffers, staged and unstaged grouping, Git Graph columns), and an agent panel accumulating operational controls (in-thread search, auto-compaction, skills management, ACP session configuration). A large share of the individual fixes carry external contributor credits.
Refine vs Zed: editorial side-by-side
Refine
0.0DEVOPS
Refine's v5 majors are out; what's shipping now is the bug tail, mostly from contributors
◆ Current state
Refine is a React framework for building internal tools and admin panels, and this window is entirely patch releases against the already-shipped v5 core and v8 MUI packages. The fixes cluster in three places: URL state parsing under syncWithLocation, access-control metadata not reaching button hooks, and CLI binary resolution inside monorepo workspaces.
◆ Where it's heading
The pattern is a framework absorbing the edge cases that only appear once a major is in real projects. Deeply nested conditional filters were being mangled because a parser depth limit was too shallow; useDeleteButton was dropping the meta prop that custom access rules depend on; refine dev broke under workspace hoisting. Notably, most of these carry different contributor names, several first-time — the maintenance load is being spread across the community rather than driven by a single team push.
◆ Prediction
Filters, access control, and router adapters have each produced multiple fixes in a single window, so expect continued patch releases in those areas; nothing in these entries points to a new major or a change in framework direction.
Zed
6.3DEVOPS
Zed ships weekly, and each week the agent panel and the Git surface get another turn of the crank.
◆ Current state
Zed runs a weekly stable release with point releases in between that exist largely to pick up newly launched models — Claude Opus 5 and the GPT 5.6 family each got a same-week patch. The substantive work splits in two: a Git surface being built toward complete in-editor staging and review (hunk-level multibuffers, staged and unstaged grouping, Git Graph columns), and an agent panel accumulating operational controls (in-thread search, auto-compaction, skills management, ACP session configuration). A large share of the individual fixes carry external contributor credits.
◆ Where it's heading
The editor is turning into an agent host as much as a text editor: ACP support for third-party agents, MCP server management, per-provider BYOK, and llama.cpp as a local provider all position Zed as neutral ground rather than one vendor's client. The patch cadence around model launches makes that posture explicit — support lands within days, and when it can't, the changelog names the blocker rather than staying quiet. In parallel, Git is being pushed toward the point where staging and review no longer require leaving the editor.
◆ Prediction
Expect the same shape to hold: a weekly feature release deepening the Git and agent panels, punctuated by out-of-band patches whenever a lab ships a model. The open question these entries raise is provider gating — Zed can only expose what model vendors unlock for third-party clients.
Alternatives to Refine and Zed
Other DevOps 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 Refine or Zed.
Workato
Workato is packaging its agents into products — IT Support and EDI Genies launched a day apart
Lokalise
Lokalise is building the measurement layer around AI translation, not just more AI translation
Okta
Okta is documenting its way into agent identity, one Cross App Access guide at a time
HashiCorp
HashiCorp is wrapping Terraform in policy and stacks for the agent-driven era.
GitHub
GitHub is hardening the registry it owns while Copilot absorbs every new model.
Recent activity from Refine and Zed
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
- 1d agoZedClaude Opus 5 lands for Anthropic and Bedrock BYOK
- 6d agoZedGit panel staging groups, finder multi-select, CSV filters
- 13d agoZedHunk-level staging multibuffers and Git Graph columns
- 16d agoZedFix language server startup failures under npm v12
- 19d agoZedGPT 5.6 Sol and Terra for ChatGPT subscribers
- 19d agoZedGPT 5.6 support in the OpenAI provider
- 3mo agoRefineCustom getList response fields survive into useList and useTable
- 4mo agoRefineMUI buttons respect a custom startIcon when text is hidden
- 4mo agoRefineCLI resolves its binary from cwd, fixing monorepo workspaces
- 4mo agoRefineDelete button forwards meta to access-control checks
- 4mo agoRefine@refinedev/[email protected]
- 4mo agoRefineDeeply nested URL filters parse correctly again
Frequently asked questions
What is the difference between Refine and Zed?
They serve adjacent needs but don't currently overlap on shipped themes. Zed is currently shipping more aggressively (velocity 6.3 vs 0.0), 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.
Is Refine better than Zed?
Sparkpulse doesn't pick a winner — we score release velocity, not feature parity. Zed is currently shipping more aggressively (velocity 6.3 vs 0.0), with 0 editorial sparks in the last 30 days against 0. For your specific use case, the alternatives sections above list other DevOps products to evaluate alongside.
What are the best alternatives to Refine?
Top Refine alternatives in DevOps are ranked by recent ship velocity. Browse the "Refine alternatives" section above for the current picks, or visit /alternatives/refine for the full list with editorial commentary on each.
What are the best alternatives to Zed?
Top Zed alternatives in DevOps are ranked by recent ship velocity. Browse the "Zed alternatives" section above for the current picks, or visit /alternatives/zed for the full list with editorial commentary on each.