FARE-20B
- Family
- FARE
- Released
- 2025-10-20
- Context
- 128K
- Parameters
- 20.9B
- Architecture
- Decoder Only
- Openness
- Open weights
- License
- CC-BY-NC-4.0Not permitted
- Weights
- Available
- Code
- Unknown
- Training
- Fine-tuned
Ethical AI solutions for customer empowerment
No tracked provider pricing is available yet.
Ethical AI solutions for customer empowerment
No tracked provider pricing is available yet.
About
FARE-20B is a multi-task evaluator model from Salesforce AI Research, fine-tuned from OpenAI's gpt-oss-20b and introduced in October 2025. It is trained with rejection-sampling fine-tuning on a large multi-domain dataset to act as a judge: comparing two responses, finding the first wrong step in a solution, verifying answers with or without a reference, and giving single ratings, using the prompt templates published with the model. Weights are on Hugging Face under CC BY-NC 4.0 (non-commercial). Fireworks offers it as an on-demand deployment with 131K context.
Providers(1)
Compare API pricing across 1 tracked provider route for input and output tokens, batch, and cached reads when available.
| Provider | Input / 1M | Output / 1M | Route |
|---|---|---|---|
| Fireworks AI | - | - | ProvisionedPartial |
Available via routers & gateways(1)
Capabilities
No model capability flags are currently sourced.
Benchmark peer barsfor Long context
No task-mapped benchmark peers are available for this model yet.
Migration checks
No linked migration route is available for this model yet.
API versions
FARE-20BSalesforce/FARE-20Bfare-20b