Fara-7B model card (Hugging Face) 24 Nov 2025
- WebVoyager 73.5
- Online-M2W 34.1
- DeepShop 26.2
- WebTailBench 38.4
Fara-7B
Fara-7B ranks first on AgentHarm.
8
Benchmarks with a score
1
Ranked first, out of five or more models
1
In the top three
Microsoft Nov 2025 Fara-7B
Open weights Small
No Epoch Capabilities Index score yet
- ScreenSpot-v2 89.3% #11 /14
- WebVoyager 73.5% #10 /11
- ScreenSpot 86.7% #8 /11
Where it ranks
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
Agents and tools
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| ScreenSpot | 86.7% | #8 of 11 | Model card | |
| ScreenSpot-v2 | 89.3% | #11 of 14 | Model card | |
| WebVoyager | 73.5% | #10 of 11 | Model card | |
| Online-Mind2Web | 34.1% | #8 of 8 | Model card | |
| WebTailBench | 38.4% | #3 of 3 | Model card | |
| DeepShop | 26.2% | #2 of 2 | Model card |
Safety and security
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| AgentHarm | 94.2% | #1 of 6 | Model card | |
| WebTailBench-Refusals | 81.9% | #2 of 2 | Model card |
Its launch post
The benchmarks the lab chose to headline, with the scores it gave.
Its system card
Every result in the longer documents the lab published with it. 4 of them aren't in the launch post.
Capabilities
- WebVoyager 73.5
- Online-Mind2Web 34.1
- DeepShop 26.2
- WebTailBench 38.4
- ScreenSpot-v2 89.3
- ScreenSpot 86.7
Safety
- AgentHarm 94.2
- WebTailBench-Refusals 81.9