Submit a run

You run the open engine against your own model and submit a signed result bundle. Deterministic results are independently re-runnable, so the leaderboard is trustworthy without us hosting any GPUs.

1. Serve a model

Any OpenAI-compatible endpoint works. For local GGUFs:

./serve/serve.sh <model-name>

2. Run the suite and produce a bundle

python -m engine.runner --models <model-name> --bundle
# -> results/<stamp>/bundle.json  (schema-valid, content-addressed, ed25519-signed)

3. Submit it

curl -X POST http://api:8000/submissions \
  -H 'content-type: application/json' \
  --data @results/<stamp>/bundle.json

What gets recorded

• Exact model identity — HF repo + revision, file SHA-256, quant, engine version.
• Sampling + serve flags + the hardware/driver it ran on (the VRAM facet).
• Per-challenge content hashes, transcripts, and scores.

The whole bundle is content-addressed and signed by your key — the signature is your root identity. Link that key to GitHub with peakstone login: optional, but it attributes runs to your handle and lets your reproductions count toward the community-verified (ranked) tier.