ELEMNT vs Felt
Felt is what teams pick for a workshop. ELEMNT is what they keep when the map becomes a system of record.
TL;DR
- Choose Felt if your team of 5 to 15 people needs to start mapping together this week, values the best collaborative map-authoring UX in the market today, and does not yet need data versioning or an audit trail.
- Choose ELEMNT if your geo dataset is a production asset that needs an audit trail, you want to know who changed what layer and when, or you are integrating map serving into a developer-built product via PMTiles and OGC API endpoints. ELEMNT is free during early access.
- Both are fine if you need file uploads, warehouse connectivity, and a professional self-hosted option. Both platforms support those workflows. The difference is what happens to the data after it lands: Felt optimizes for collaborative map-making, ELEMNT optimizes for data integrity and open-format serving.
Comparison table
| Feature | Felt | ELEMNT |
|---|---|---|
| Pricing model | Annual-only, seat-based (10-seat minimum on Professional), quote-only | Free during early access |
| Free tier | Personal free (non-commercial, no data file upload) | Free during early access |
| Self-host / on-prem | Felt Server (1 instance on Professional, unlimited on Enterprise) | On the roadmap |
| REST API + SDK | REST API + JS SDK on Enterprise | OGC API surface at GA; CLI and SDKs on the roadmap |
| MCP support | Felt AI + MCP on Enterprise | MCP server on the roadmap |
| Data versioning | None on any plan (re-verified June 12, 2026) | Git-style versioning (Q4 2026 roadmap) |
| Real-time collab | Best in market today | Q1 2027 roadmap |
| OGC API exposure | Imports WMS/WMTS; does not expose OGC API endpoints for your data | OGC API Features + Style JSON |
| Open serving formats | Export to standard formats on all plans; serving substrate is not open-format based | PMTiles, GeoParquet at the storage layer at GA (COG, Zarr roadmap) |
| Warehouse connectivity | Postgres, Snowflake, BigQuery, Databricks, S3 (Enterprise live sources) | On the roadmap: BigQuery, MongoDB, Firestore, ClickHouse, PostGIS, Snowflake (import via open formats at GA) |
| AI map authoring | Felt AI (natural language to map, all plans) | None; the planned MCP server (roadmap) is data-platform-facing |
| SOC 2 Type II | Both Professional and Enterprise | In progress, not yet certified |
| STAC support | None | Roadmap (post-GA) |
| Offline field app | Beta | None |
Source: felt.com/pricing, re-verified June 12, 2026.
The structural difference
Felt and ELEMNT share a market segment but are built around different assumptions about what a mapping platform is for.
Felt's founding insight is that maps are documents that teams collaborate on. The product is optimized for the moment when five people in a planning meeting all need to look at the same geographic picture, annotate it, argue about it, and leave with a shared artifact. The onboarding is fast. The UX is genuinely good. Felt AI turns natural-language descriptions into map layers. The collaborative editing is real-time and polished. For that workflow, Felt is the right tool today and the gap between Felt and any competitor on this dimension is real. If you are choosing a platform for a team that starts a new map every week, Felt has earned its position.
ELEMNT's founding insight is different: the map layer inside a production system is a dataset, not a document. When a logistics SaaS serves route overlays to 200,000 users, or a national mapping agency publishes infrastructure boundaries that feed regulatory decisions, the question is not “can my team annotate this in a browser?” The questions are: who changed this layer, when, and can we roll it back? Can the compliance auditor see the diff? Can we branch the dataset for a new project without duplicating 40 GB of geometry? Can every consumer of this data, regardless of their GIS stack, pull it via a standards-compliant OGC API endpoint?
Those are data integrity questions, not collaboration UX questions. They require a different architecture.
On data serving, the difference is at the storage layer. Felt's substrate is an application database that serves its own rendering pipeline. Felt lets you export your data in standard formats (Shapefile, GeoJSON, GeoPackage) on every plan, so you are not locked in at the export layer. But the data is not natively served as PMTiles, COG, or GeoParquet from object storage. ELEMNT's serving layer is cloud-object-native: PMTiles served direct from S3, GCS, or your own CDN, plus GeoParquet for analytical queries. COG for raster and Zarr for multidimensional data are on the post-GA roadmap. No tile server sits between your object store and the consumer. You can take those files and serve them yourself, from a different platform, or directly via a CDN, with zero vendor lock-in at the format layer.
On OGC API exposure, Felt can import WMS and WMTS sources. It does not expose OGC API endpoints for the data you upload. If a downstream system needs to query your dataset as an OGC API Features service, Felt does not provide that endpoint. ELEMNT exposes your data via OGC API Features and Style JSON natively. For public-sector procurement that requires OGC API compliance, that distinction is a hard requirement, not a preference.
On versioning, the gap is total. Felt has no history, no snapshots, no branches, no time-travel, and no rollback on any plan (re-verified felt.com/pricing, June 12, 2026). ELEMNT ships Git-style versioning in Q4 2026. If your geo dataset needs an audit trail, the choice is straightforward.
The pricing model difference is structural too. Felt prices by seat, annually, with a 10-seat minimum on Professional. If you have a team of three developers building a product, you are paying for 10 seats you do not use. ELEMNT is self-serve and free during early access, with no contract required to start. Public pricing is not published yet; when it is, it will be on this site first.
Pricing and cost
Felt pricing
Felt does not publish list prices. According to Felt's own FAQ, prices are not public. The Professional plan includes 10 full seats, unlimited viewers, 25 GB of processing per month, 100K map views per month, and one Felt Server instance. Enterprise adds live cloud sources (Postgres, Snowflake, BigQuery, Databricks, S3), advanced SQL, raster streaming, AI and MCP, REST API and JS SDK, SSO and SAML, regional hosting with data residency and VPC, and unlimited Felt Server instances. Both plans are annual-only (confirmed by Felt's own FAQ: “Can I pay monthly?”). Felt does not auto-bill overages. If usage exceeds plan limits, the Felt sales team contacts you.
A free Personal tier exists but is non-commercial and does not allow data file uploads.
ELEMNT pricing
ELEMNT is free during early access. No credit card, no quote calls, no hidden tiers.
ELEMNT is free during early access: self-serve sign-up with no sales conversation required. Public pricing is not published yet; when it is, it will be on this site first. ELEMNT does not offer geocoding; pair a third-party geocoder if you need one.
When Felt is the right choice
Felt is worth saying clearly: for collaborative map authoring, Felt is the best product in the market today. That is not a token concession. It is a factual assessment of where the UX is right now.
If your primary workflow is a planning team sitting down to annotate a map together, run spatial analysis on uploaded data, and share a live map link with stakeholders, Felt's onboarding, real-time editing, and Felt AI are polished in a way that ELEMNT's Q1 2027 collaboration roadmap cannot match today. You would be waiting six months for a feature Felt ships now.
Felt AI is a genuine capability. Natural-language map authoring, where you describe what you want and Felt builds the layer, is live and in production. ELEMNT has no equivalent today; the planned MCP server (on the roadmap) will connect AI agents to the data platform layer, not to map authoring. These are different workflows.
If your team needs to be productive on Monday morning with zero configuration, Felt wins the trial experience. The onboarding does not require a developer.
For teams that need SSO, SOC 2 Type II compliance, and enterprise security today, Felt ships both on Professional and Enterprise. ELEMNT's SOC 2 certification is in progress and was not complete as of June 2026.
If you have a team of 10 or more people who will use the platform heavily for collaborative map-making and you have an annual budget for it, the seat-based model at scale may work out to a reasonable per-user cost. Ask Felt for a quote and compare it against ELEMNT's usage projection from the calculator.
Migration and coexistence path
Felt and ELEMNT can coexist in a workflow, and for some teams that is the right answer.
A common pattern: use Felt for the collaborative layer (team workshops, stakeholder annotation, planning sessions) and ELEMNT for the data platform layer (versioned datasets, OGC API serving, developer-built map products). The two are not architecturally incompatible. Felt can consume OGC API services as WMS/WMTS sources, so ELEMNT-served data can flow into a Felt map.
For teams moving their production data layer to ELEMNT, the migration path depends on where the data currently lives in Felt. Felt exports standard formats (Shapefile, GeoJSON, GeoPackage, CSV) on all plans. That export is the migration input. From there:
- Export your Felt datasets in GeoJSON or GeoPackage format.
- Ingest into ELEMNT (direct upload, or connect your source datastore if the data originates there).
- Configure PMTiles serving and OGC API endpoints.
- Validate the serving path end to end before switching production traffic.
- If you are migrating tile serving for a developer-built frontend, the MapLibre GL JS code that consumes ELEMNT's PMTiles endpoint is the same as any PMTiles-compatible source.
Felt's collaborative UX is not something ELEMNT replaces today. If your team uses Felt for map workshops and you are adding ELEMNT for data versioning and OGC serving, keeping both in the stack is a reasonable choice until ELEMNT's Q1 2027 collaboration layer ships.
FAQ
- Does Felt have version control for data?
- Felt has no data versioning on any plan. There is no history log, no snapshot, no branch, no rollback, and no time-travel query for uploaded datasets. This was re-verified against felt.com/pricing on June 12, 2026: the words “history”, “snapshot”, “version”, and “rollback” do not appear in the pricing or features documentation. For a dataset that a single user annotates occasionally, this is not a problem. For a dataset that feeds regulatory decisions, gets edited by multiple departments, and needs an audit trail proving what the boundary looked like on a specific date, the absence of versioning is a structural gap. ELEMNT ships Git-style versioning (branches, diffs, merge, audit log) in Q4 2026. (Source: felt.com/pricing, accessed June 12, 2026.)
- Does Felt have an API or self-hosting option?
- Yes on both. Felt ships a REST API and a JavaScript SDK on Enterprise, and an MCP server for AI agent integration. Felt Server (self-hosted) is available on Professional (one instance) and Enterprise (unlimited instances). These are production capabilities that Felt has shipped. The architectural difference from ELEMNT is not presence or absence of these features but where the platform sits: Felt is an application platform with API and self-host options layered on top. ELEMNT is a data platform where the serving substrate (PMTiles and GeoParquet from object storage, with COG and Zarr on the roadmap) is open-format native from the ground up, and the OGC API endpoint is the primary interface, not an add-on. (Source: felt.com/pricing, accessed June 12, 2026.)
- How does Felt's pricing model differ from ELEMNT's?
- Felt prices by seat on an annual contract with a 10-seat minimum on Professional. The list price is not public; Felt's FAQ confirms this deliberately. ELEMNT is self-serve, with no contract required to start, and free during early access. Public pricing is not published yet; when it is, it will be on this site first. The model difference matters when your team is small (fewer than 10 people, where Felt's 10-seat floor means paying for unused seats) or when you want to start without a procurement cycle. (Source: felt.com/pricing, accessed June 12, 2026.)
- Is ELEMNT collaborative like Felt?
- No, not today. Real-time collaborative map editing is on ELEMNT's roadmap for Q1 2027. Until that ships, Felt has the better collaborative UX. If your team's primary need is to annotate and discuss maps together in real time, Felt wins that comparison today and will continue to for at least the next two quarters. ELEMNT's collaboration investment is real (Q1 2027 is a committed roadmap quarter, not a vague future plan), but committing to a platform based on unshipped features is not the right advice. If you need real-time collaboration now, use Felt. If you need data versioning and open-format serving now, and can wait on collaborative UX, ELEMNT is the choice. If you need both today, the honest answer is that no single platform ships all of it yet.
- Which platform uses open formats under the hood?
- ELEMNT's serving substrate is open-format native. PMTiles (an open, single-file archive format for map tiles) served direct from object storage means no proprietary tile server and no vendor-specific query format sitting between your data and your frontend. GeoParquet for analytical queries completes the GA stack, with COG (Cloud Optimized GeoTIFF) for raster and Zarr for multidimensional arrays on the post-GA roadmap. Every format is openly specified. You can read, serve, and migrate these files with standard open-source tools. Felt takes a different approach: export is available in standard formats (Shapefile, GeoJSON, GeoPackage) on all plans, which means you can always get your data out. But the internal serving substrate is not open-format based, and Felt does not expose OGC API endpoints for your uploaded data. The distinction matters most when downstream systems need to query your dataset via a standard OGC API interface, or when you want to serve tiles directly from your own object store without routing through a managed service. (Source: felt.com/pricing, accessed June 12, 2026.)
Start free
Start free during early access. No credit card, no quote calls, no hidden tiers. Import your first dataset and serve your first tiles.