Search Intelligence Score
63.6
+35.5 lift from search
050100
- Lift from search
- +35.5
- 28.1 without search
- Cost per 1K tasks
- $43.4
- Below the median score of 64.3, so not ranked on cost
Accuracy by benchmark
Accuracy on each eval, with search (solid) and without (light).
- DSQA81.8 vs 41.4
- HLE41.0 vs 33.0
- WISER68.0 vs 10.0
Cost per 1K tasks
How the cost splits between model inference, search calls, and page extraction.
- Inference
- $22.7
- 52%
- Search
- $5.85
- 13%
- Extract
- $14.9
- 34%
Compared with
Other models from the same provider alongside models that score closest.
- Mercury 363.6Cost $43.4 · Lift +35.5
- Kimi K364.2Cost $266 · Lift +33.6
- Haiku 5.562.9Cost $24.3 · Lift +39.4
- MiMo v2.6 Pro64.4Cost $76.5 · Lift +34.7
- Mistral Large 464.4Cost $136 · Lift +43.0
| Model | Score | Lift | Cost per 1K tasks |
|---|---|---|---|
| Mercury 3 | 63.6 | +35.5 | $43.4 |
| Kimi K3Kimi K3 | 64.2 | +33.6 | $266 |
| Haiku 5.5Claude Haiku 5.5 | 62.9 | +39.4 | $24.3 |
| MiMo v2.6 ProMiMo v2.6 Pro | 64.4 | +34.7 | $76.5 |
| Mistral Large 4Mistral Large 4.0 | 64.4 | +43.0 | $136 |
Search Intelligence Score averages accuracy on DSQA, HLE, and WISER with equal weight; cost is averaged the same way, with cost shown per 1,000 tasks. Evaluated . Leaderboard last updated . Full methodology.
# Mercury 3 · Search Capability Leaderboard
- Search Intelligence Score: 63.6 (#25 in Search Intelligence)
- Without search: 28.1 · Lift from search +35.5
- Cost per 1K tasks: $43.4 (below the median score of 64.3, so not ranked on cost)
- Evaluated:
- Lab: Inception Labs · Weights: Closed · Model id: inception/mercury-3
## Accuracy by benchmark
| Suite | With search | Without search |
|---|---|---|
| DSQA | 81.8 | 41.4 |
| HLE | 41.0 | 33.0 |
| WISER | 68.0 | 10.0 |
## Compared with
| Model | Score | Lift | Cost per 1K tasks |
|---|---|---|---|
| Mercury 3 | 63.6 | +35.5 | $43.4 |
| Kimi K3 | 64.2 | +33.6 | $266 |
| Claude Haiku 5.5 | 62.9 | +39.4 | $24.3 |
| MiMo v2.6 Pro | 64.4 | +34.7 | $76.5 |
| Mistral Large 4.0 | 64.4 | +43.0 | $136 |
Leaderboard last updated . Methodology: /leaderboard#methodology