Anthropic
Claude 4.5 Sonnet
ReasoningClosed weights
- Intelligence
- 19.3
- Rank 257 of 744 measured models
- Coding
- 52.1
- Rank 115 of 298 measured models
- Output speed
- 41 t/s
- Rank 360 of 403 measured models
- Price/1M
- $6
- Rank 587 of 659 measured models
- Context window
- 1M
- Rank 84 of 1199 measured models
Claude 4.5 Sonnet (Non-reasoning) scores 19.3 on the Artificial Analysis Intelligence Index, ranking 257th of 744 measured models. The closest models on this index are Step 3.7 Flash (19.5), Qwen3.5 35B A3B (19.3), and LongCat 2.0 (19.1). At $6 per million tokens (3:1 input/output blended price), it is more expensive than Step 3.7 Flash ($0.438), Qwen3.5 35B A3B ($0.688), and LongCat 2.0 ($1.30). With an output speed of 41 t/s, it is slower than Step 3.7 Flash (185 t/s), Qwen3.5 35B A3B (147 t/s), and LongCat 2.0 (67.2 t/s).
Additional benchmarks11
Design Arena · 3d51.0%
Design Arena · Asciiart56.1%
Design Arena · Codecategories51.5%
Design Arena · Dataviz47.4%
Design Arena · Fullstack43.7%
Design Arena · Gamedev51.1%
Design Arena · Mobileapps52.9%
Design Arena · Svg52.2%
Design Arena · Uicomponent49.5%
Design Arena · Webapps43.2%
Design Arena · Website51.8%
Models at a similar level
The closest models on the Intelligence Index, one configuration per family.
- K-EXAONE 2.0 0803
LG AI Research · — · —
19.7+0.4——Compare - Step 3.7 Flash
StepFun · $0.438 · 185 t/s
19.5+0.2$0.438185 t/sCompare - Qwen3.5 35B A3B
Alibaba · $0.688 · 147 t/s
19.3$0.688147 t/sCompare - LongCat 2.0
LongCat · $1.30 · 67.2 t/s
19.1−0.2$1.3067.2 t/sCompare
