Bearing
Find the right AI model for the job.
Bearing helps you choose an AI model for a particular task. Explain what you need, see a recommendation with meaningful alternatives, and understand the evidence behind it.
Try Bearing · Start with the user guide
What makes it different?
- Task first: recommendations start with what you're trying to achieve.
- Explainable: inspect trade-offs such as quality, cost, privacy, openness and sustainability.
- Practical: run and compare supported models, including local options.
- Open to challenge: assumptions, evidence and the methodology can all be questioned.
A ranking score is not a probability of success. Bearing keeps model claims, benchmarks, current availability and outcomes distinct.
Find your way
| I want to… | Read |
|---|---|
| Use Bearing | User guide |
| Understand recommendations | How Bearing works |
| Examine the scoring and evidence | How we rate models |
| Explore the published data | Public data model |
| Correct data or question a result | Contribute |
| Work on the code or database | Developer contribution guide |
Help shape Bearing
You don't need technical skills or database access. Report a factual error, share evidence, explain a confusing recommendation, or propose a better way to decide. Start with the contribution routes.
Bearing is built by The Good Ship.