Search Capability Leaderboard
Claude Opus 5
Anthropic
#4 in Search Intelligence · #13 in Search Efficiency
Search Intelligence Score
70.0
+32.8 lift from search
050100
- Lift from search
- +32.8
- 37.2 without search
- Cost per 1K tasks
- $1,012
- #13 in Search Efficiency
- Time per task
- 342s
- Search plus inference time to complete one task
Accuracy by benchmark
Accuracy on each eval, with search (solid) and without (light).
- DSQA81.9 vs 48.5
- HLE54.0 vs 48.0
- WISER74.0 vs 15.0
Cost per 1K tasks
How the cost splits between model inference, search calls, and page extraction.
- Inference
- $996
- 98%
- Search
- $4.18
- 0%
- Extract
- $11.2
- 1%
Compared with
Other models from the same provider alongside models that score closest.
- Opus 570.0Cost $1,012 · Lift +32.8 · 342s
- Fable 5.172.7Cost $1,655 · Lift +30.3 · 512s
- Sonnet 566.6Cost $688 · Lift +43.8 · 846s
- Opus 5.575.4Cost $949 · Lift +31.2 · 512s
- GPT-6 Astra70.8Cost $401 · Lift +27.9 · 83.7s
| Model | Score | Lift | Cost per 1K tasks | Time per task |
|---|---|---|---|---|
| Claude Opus 5 | 70.0 | +32.8 | $1,012 | 342s |
| Fable 5.1Claude Fable 5.1 | 72.7 | +30.3 | $1,655 | 512s |
| Sonnet 5Claude Sonnet 5 | 66.6 | +43.8 | $688 | 846s |
| Opus 5.5Claude Opus 5.5 | 75.4 | +31.2 | $949 | 512s |
| GPT-6 AstraGPT-6 Astra | 70.8 | +27.9 | $401 | 83.7s |
Search Intelligence Score averages accuracy on DSQA, HLE, and WISER with equal weight; cost and time are averaged the same way, with cost shown per 1,000 tasks. Latest update September 25, 2026. Full methodology.
# Claude Opus 5 · Search Capability Leaderboard
- Search Intelligence Score: 70.0 (#4 in Search Intelligence)
- Without search: 37.2 · Lift from search +32.8
- Cost per 1K tasks: $1,012 (#13 in Search Efficiency)
- Time per task: 342s
- Lab: Anthropic · Model id: claude-opus-5
## Accuracy by benchmark
| Suite | With search | Without search |
|---|---|---|
| DSQA | 81.9 | 48.5 |
| HLE | 54.0 | 48.0 |
| WISER | 74.0 | 15.0 |
## Compared with
| Model | Score | Lift | Cost per 1K tasks | Time per task |
|---|---|---|---|---|
| Claude Opus 5 | 70.0 | +32.8 | $1,012 | 342s |
| Claude Fable 5.1 | 72.7 | +30.3 | $1,655 | 512s |
| Claude Sonnet 5 | 66.6 | +43.8 | $688 | 846s |
| Claude Opus 5.5 | 75.4 | +31.2 | $949 | 512s |
| GPT-6 Astra | 70.8 | +27.9 | $401 | 83.7s |
Latest update September 25, 2026. Methodology: /leaderboard#methodology