Platform
Data, models and evaluation.
Behind the roster is the work that makes it dependable: curated expert data, specialized training, finetunes and the evaluation harness that tests each release.
Routing
One API. Seven models. Predictable behavior.
Every request goes to the same OpenAI-compatible endpoint and is served by the model you name. Each model publishes its architecture and scale, and the model you call is the model that answers.
Curated data
Expert-authored corpora, selected for signal.
We train on material written by leading practitioners in science, engineering, computer science, cryptography and economics. Sources are screened to exclude low-signal web content and institutionally captured material. Provenance is recorded at ingest, so every passage can be traced to its origin.
Training
The knowledge lives in the weights.
Each model is specialized on the curated expert corpus, so what it knows is learned during training and carried in its weights. The provenance of every source is recorded at ingest, and each answer states its claims, their sources and its uncertainty.
Finetuning
Your data. Your model.
We specialize a model on your material and serve it through the same API under its own name. It joins your roster with its architecture, data provenance and evaluation record documented.
Evaluation
Tested before release.
Every model and finetune is evaluated against task-specific harnesses: source-trace accuracy, claim support, calibration and factual accuracy. The harness is designed to find failure before users do.
API
One key. Every model.
The endpoint is OpenAI-compatible, so existing SDKs, tools and agents work without changes. Choose a model by name and send a request.
curl https://api.trybased.ai/v1/chat/completions \
-H "Authorization: Bearer $BASED_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "based",
"messages": [{ "role": "user", "content": "Show your work." }]
}'Ways to work with us.
API access
Request a key, point your existing client at the endpoint and choose a model.
Custom build
A curated dataset or a finetune, scoped in writing and delivered into your stack.
Research partnership
Ongoing model development for teams that need a research partner on call.
Frequently asked questions.
Who builds your models?
Based AI builds every model on the roster. Each one starts from an open-weight foundation, is specialized on curated, expert-authored data and is evaluated before release. We publish the architecture and parameter scale of every model.
What data do you train on?
Curated corpora authored by leading experts in science, engineering, computer science, cryptography, economics and metaphysics. Sources are selected for signal and screened to exclude low-signal web content and institutionally captured material. Provenance is recorded for every source.
What do you disclose about each model?
Architecture, parameter scale, intended use, training-data provenance and evaluation record. Together they tell you what a model can handle, what it costs to run and where it is likely to fail.
Do you train on our data?
No. Client data is scoped, isolated and used only for the engagement it was provided for. A finetune trained on your material is served under its own name, for your use only.
What does an engagement look like?
You send us a short description of the work. We reply with a scope, a timeline and a fixed price. You work directly with the people building your system.
Can my agent use the API?
Yes. The endpoint is OpenAI-compatible, so agents and SDKs that support that format can call any model on the roster by name.
V — Access
Build with Based AI.
API access, custom builds and finetunes start with a short request. Tell us what you are building and a member of our team will reply.