Compared with SnapGene

Every claim about SnapGene on this page comes from SnapGene's own pages, retrieved 2026-08-08, and is recorded once, in the open claim register, with its source and date. Nothing here is from a comparison blog or from memory — several roundups were read during the research and they disagreed with each other, which is why they are disqualified as sources.

If you find a row that is out of date, it is a bug in this page. Vendors change their products, and a comparison that is not dated cannot be checked.

The short versionCopy Anchor Link

SnapGene is the plasmid editor most labs actually use, and it is good at what it does. The two differences that matter are where it runs and where the free line falls.

That last row is the honest headline. SnapGene's free Viewer mode is a viewer: it opens maps and traces, annotates and shares. Simulating a cloning reaction, editing a sequence and aligning anything are in the paid product. Here, every one of those is free, and always will be — the pricing rule in this project is that local scientific correctness is never paywalled.

Where SnapGene is aheadCopy Anchor Link

Stated as plainly as the reverse, because a comparison that only flatters us is marketing.

• More cloning methods, by name. SnapGene lists Restriction, Gateway, Gibson, NEBuilder HiFi, In-Fusion, TA & GC, TOPO and Golden Gate. The workbench ships Gibson, Golden Gate, grammar assembly (MoClo/GoldenBraid), restriction digest, and Gateway BP/LR and Cre-lox recombination (Recombination) as surfaces; TOPO and TA/GC cloning have none, and NEBuilder HiFi / In-Fusion are homology-based reactions that Gibson's engine covers mechanically without being named as their own workflow. If your protocol is written around a named kit, SnapGene names it and we do not.
• De novo Sanger contig assembly with CAP3. SnapGene assembles Sanger reads into contigs. We do not: our alignment engine is partial-order alignment for multiple sequence alignment, plus consensus and mutation calling against a reference. Assembling reads into a novel contig is a different operation and we do not have it.
• A feature-annotation library built over many years. Ours searches six sourced corpora (plasmid parts, resistance genes, fluorescent proteins, Swiss-Prot plasmid proteins, curated GenBank elements and Rfam), each hit citing the record it came from; see Auto-Annotate. SnapGene's library is its own and we have not measured one against the other.
• It is the format everyone shares. .dna files are the lingua franca of plasmid sharing. We read and write them, including the cloning-history tree — but SnapGene is the reason the format matters, and a format's centre of gravity is worth more than any one reader of it.

Where this is aheadCopy Anchor Link

• Nothing to install, and nothing uploaded. It opens in a browser tab, on any machine, including one you do not administer. There is no server in production, so your sequences physically cannot go anywhere.
• The simulation half is free. See the table above.
• The methods are checkable. Every scientific capability is checked against published implementations — pydna, DnaCauldron, QUEEN, DnaChisel, SeqTrace, sangerseqR, poapy, crisprScore and others — asserted byte-for-byte against the real library on the same input. Where a number is approximate or a model is missing, the tool's page says so. We do not know what SnapGene's algorithms are, because they are not published, and we make no claim about whether they are right — that is a real difference in kind, not a claim about quality.
• More than one algorithm where the field has more than one. Melting temperature offers eight published models because they disagree by up to 21 °C on the same 20-mer, and which one you want depends on your polymerase and buffer. Dot plots offer three. Codon optimization offers four objectives.
• File formats beyond the plasmid world. Indexed BAM and VCF read by locus rather than ingested, plus FASTQ, trees, MSA formats, Parquet and Excel. SnapGene's published feature list does not include these.
• Python and R in the same tab. Pyodide and WebR with the project's files mounted, so your own analysis runs on your own data without exporting it anywhere.
• The whole thing is scriptable. Every engine is a pure headless function you can call directly, and there is a workflow DAG for chaining them.
• The construction history comes back OUT. A .dna file's cloning history is read by walking the reaction graph the file embeds — each step, its inputs and its parameters — and it can then be written out as a GenBank whose own annotations state how the construct was made, an OpenCloning strategy document that loads in pydna, a runnable Python script that rebuilds the construct, or a numbered methods section. The same reading works on a QUEEN-authored GenBank and on an OpenCloning strategy. That is the part worth knowing before you decide anything below: the record of how a construct was made is normally the thing that does not travel between programs, and here it leaves in formats that are not ours. See Construction history.

What is genuinely similarCopy Anchor Link

Both do: plasmid maps with feature annotation, restriction digest with a simulated agarose gel and MW markers, primer design, PCR and mutagenesis simulation, Gibson and Golden Gate, CRISPR guide design, enzyme sets, sequence alignment, and a graphical record of how a construct was made. On those, choose on interface and price rather than capability.

Which to useCopy Anchor Link

Honestly, and this is the useful part:

• You have a SnapGene licence and it does what you need → keep using it. This is not a migration argument.
• You do not have a licence, and Viewer mode will not simulate your cloning → that is the case this project was built for.
• You cannot install software — a shared machine, a locked-down institution, a Chromebook, a phone → browser-only is the whole point.
• Your data cannot leave the machine — patient-derived sequence, unpublished work, an institution that forbids uploads → nothing here is uploaded, by construction.
• You need to script it, or check the algorithm → the engines are callable and their provenance is published.
• You need de novo Sanger assembly, or a named kit workflow we do not have → SnapGene, today.
• You want your construction records in an open format, whatever you use next → open the files here and export the history as GenBank annotations, an OpenCloning document or a script. This is worth doing even if you go on using SnapGene for everything else, and it is the one thing on this page that costs nothing to try: it is reading your own files and writing them back out.

SourcesCopy Anchor Link

No price is quoted on this page. SnapGene does not publish a single figure — it varies by licence type, region and academic status — and every number found in third-party comparisons disagreed with the others. A sourced-looking wrong price would be worse than none. Ask the vendor for a quote.

RelatedCopy Anchor Link

The other comparisons: the comparison index · Geneious Prime · Benchling

The tools this page mentions: Gibson Assembly · Golden Gate · Restriction Digest · Auto-Annotate · Melting Temperature · Variant Calls