# JapanFold

> JapanFold is a hosted HTTP API for biomolecular structure prediction (Boltz-2, ESMFold-2, OpenFold3, OpenBind-0, RoseTTAFold3, OpenDDE and AF2-IG), binder design (BoltzGen, RFdiffusion3 and PXDesign) and protein embeddings (ESMC and SaProt), run on Tenstorrent hardware in Japan. Nothing to install and no local GPU.

- Base URL: https://api.japanfold.aiand.com
- Auth: every call except `GET /v1/models` sends `Authorization: Bearer $JAPANFOLD_API_KEY`. Keys look like `jf_live_...`; the convention is to keep yours in the `JAPANFOLD_API_KEY` environment variable.
- No key yet: sign in at https://japanfold.aiand.com/account/token with an email address. A new account starts with $100.00 of credit.
- Every call is an async job: submit, then `GET /v1/jobs/{id}` (add `Prefer: wait=60` to block until it finishes), then `GET /v1/jobs/{id}/results` for the scores. A succeeded job read by its id carries `download`, one shell command that fetches every result file.
- MCP: the same API is a remote MCP server at https://api.japanfold.aiand.com/mcp with the same key. Follow a job with `get_job` and `wait=100`: one call holds up to 100 s.

Fold a sequence:

```bash
curl -s https://api.japanfold.aiand.com/v1/predictions \
  -H "Authorization: Bearer $JAPANFOLD_API_KEY" -H 'Content-Type: application/json' \
  -d '{"model":"esmfold2","sequence":"MKTAYIAKQRQISFVKSHFSRQLEERLGLIEVQ"}'
# -> {"id": "...", "status": "queued", ...}
curl -s -H "Authorization: Bearer $JAPANFOLD_API_KEY" -H 'Prefer: wait=60' https://api.japanfold.aiand.com/v1/jobs/<id>
curl -s -H "Authorization: Bearer $JAPANFOLD_API_KEY" https://api.japanfold.aiand.com/v1/jobs/<id>/results
```

## Docs

- [Agent skill](https://japanfold.aiand.com/skill.md): the full guide for an agent, every model and input shape
- [OpenAPI 3.1 contract](https://api.japanfold.aiand.com/v1/openapi.json): the source of truth for every request and response
- [Models and limits](https://api.japanfold.aiand.com/v1/models): the live catalog, no key needed
- [Documentation](https://japanfold.aiand.com/docs/): authentication, predictions, designs, embeddings, jobs, errors
- [Full docs as one text file](https://japanfold.aiand.com/llms-full.txt)