Pay As You Go
Most popularFor individual researchers
- Credits never expire
- Files up to 20GB
- Email support
Enterprise
For core facilities
- Unlimited conversions
- No file size limit
- Dedicated support
How credits work
Each conversion costs credits based on file size. Larger files require more compute time and memory.
| File size | Credits per conversion | Pay As You Go cost |
|---|---|---|
| < 2 GB | 1 credit | $2.50 |
| 2 GB - 5 GB | 2 credits | $5.00 |
| 5 GB - 10 GB | 3 credits | $7.50 |
| 10 GB - 20 GB | 5 credits | $12.50 |
Frequently asked questions
Can I try it before I pay?
Yes. Files under 100 MB can be converted for free.
What file formats are supported?
Seurat objects (.rds) and AnnData objects (.h5ad). We convert in both directions with full metadata preservation including cell metadata, gene metadata, embeddings, and graphs.
Do credits expire?
Pay As You Go credits never expire. Lab subscription credits reset each billing cycle.
Can I pay with a grant or PO?
Yes. For Lab and Enterprise plans we can invoice against a grant or purchase order. Contact hello@singlecellconvert.com to get set up.
Why is this conversion hard?
Seurat and AnnData were designed with fundamentally different data models. Existing tools like SeuratDisk (archived, unmaintained) and sceasy (drops metadata) fall short. Here's what makes it non-trivial:
Seurat → AnnData
- Multi-assay vs. single matrix — Seurat natively supports multiple assays (RNA, ADT, ATAC), while AnnData stores a single .X matrix with additional layers
- Dimensional reductions — Seurat stores PCA, UMAP, and tSNE in named @reductions slots that must map to .obsm
- Graph storage — Seurat's SNN/NN graphs use a different sparse matrix convention than AnnData's .obsp
AnnData → Seurat
- Ambiguous .X matrix — The main matrix could be raw counts, log-normalized data, or scaled values. We use heuristics to determine the correct Seurat slot
- Layer mapping — AnnData layers need to map to the right combination of Seurat's counts, data, and scale.data slots
- Embedding keys — AnnData uses arbitrary keys in .obsm that must be mapped to Seurat reduction objects with the correct key prefix
Ready to convert?
Files under 100 MB convert for free — no account required.
When you need this
Common scenarios where researchers need to convert between Seurat and AnnData formats.
Seurat → AnnData
- Your analysis was done in R (Seurat) but you need Python tools like Scanpy, scvi-tools, or CellRank for downstream analysis
- You want to submit data to a repository that requires .h5ad format, such as CELLxGENE Discover
- Collaborators work in Python and need your R-processed single-cell data in a native format
- You need a SeuratDisk alternative since it is archived and no longer maintained
AnnData → Seurat
- Your analysis was done in Python (Scanpy, scvi-tools) but you need R-based tools like Seurat, Monocle, or CellChat
- You downloaded data from CELLxGENE (which distributes .h5ad files) and need it in Seurat format
- Collaborators work in R and need your Python-processed data in a native format