Platform
Getting started
An account takes a minute, the CPU engines are free forever, and a first real analysis fits in an afternoon. Credits only enter the picture when you reach for a GPU.
Free
CPU engines
$25
for 250 credits
1,000
constructs to browse
65
organisms available
01
Sign up, then look around
Creating an account needs an email address and nothing else. No card, no sales call, no trial clock counting down in the corner. You land on an empty dashboard with the full catalog already available: 1,000 reference constructs across 65 organisms, 70 series and 6 tiers from Pepper up through Tier 5, spanning 100 SHU to 16,000,000 SHU.
Browsing that catalog is the best twenty minutes a new account can spend. Open an entry like SS-0421 and you see the whole record — the host, the target, the pathway, the construct, the predicted and measured heat. It doubles as a worked example of what your own projects will look like when they are finished.
02
What the free tier includes
Every CPU engine, unmetered
Flux balance analysis, codon optimization, primer design, sequence annotation, construct assembly, phylogenetics and validation. Not a sample of each — the real engines, run as often as you like.
The complete catalog
All 1,000 constructs, searchable and filterable, with the genome browser, the pathway map and the heat ladder explorer alongside. The reference layer is not gated behind a plan.
Projects and exports
Create projects, revise them, and export finished designs as GenBank, FASTA, SBOL or CSV with a full manifest. A design you make on the free tier leaves with you in open formats.
API access
Issue a key and drive the CPU engines programmatically, which is enough to build and test a complete integration before spending anything at all.
GPU engines when you need them
Variant scanning, structure prediction and docking run on GPU workers and draw on prepaid credits. $25 buys 250 credits; one credit covers roughly one Standard job.
For teams running it in production
Scoped API keys, webhook callbacks, reserved GPU capacity so scheduled batches finish predictably, and SSO through your own identity provider.
03
Your first analysis
- Minute 1
- Create an account, land on the dashboard
- Minute 5
- Open a catalog entry near your interest
- Minute 15
- New project — pick host, tissue, target
- Minute 20
- Run flux analysis — free, answers in-page
- Minute 30
- Read the ceiling, revise the target, re-run
- Hour 1
- Assemble a draft construct, validate it
- Spend so far
- Nothing
That sequence is the honest version of a first session, and the last line is the point. A real feasibility answer for a real host against a real target costs nothing, because flux modelling runs on CPU. Teams routinely use the free tier alone to kill ideas that would have consumed a greenhouse season, which is a return on an afternoon.
Credits become relevant at the moment you want to improve an enzyme rather than establish whether the host can fund one. Variant scanning and structure prediction are GPU work: choose Standard for routine scans, Performance when a shortlist matters, Max when a single structure is carrying a decision. The cost appears in the run console before you commit, never after.
04
Bring the rest of the team
Invite colleagues by email from any project. Membership is explicit and per-role, so a collaborator sees the projects you added them to and nothing else. Shared projects carry their full history, which means a new member arrives to the reasoning as well as the result rather than to a folder of files and a conversation they missed.
Credits pool at the workspace level, so a team buys once instead of person by person, and the run console attributes every GPU job to the member or key that submitted it. When the work moves from experiment to programme, enterprise adds SSO, reserved capacity and webhook delivery without anything having to be rebuilt.
Ready to run it
Try this in the Splice Platform.
CPU engines are free to run. GPU engines run on prepaid credits.
Related