Anthropic
Anthropic: Claude Sonnet 5
anthropic/claude-sonnet-5
Sonnet 5 is Anthropic's most capable Sonnet-class model, with frontier performance across coding, agents, and professional work. It supports adaptive thinking with selectable reasoning effort levels (low, medium, high, max, and x-high), a 1M-token context window, and text, image, and file inputs. Sonnet 5 uses an updated tokenizer and includes real-time cyber safeguards that block certain high-risk dual-use activities.
/ 1M tokens
/ 1M tokens
tokens
Limits
Context
1,000,000
tokens
Max output
128,000
13% of the context window
Model facts
Author
Anthropic
Tokenizer
Claude
Modality
Text + Image + File -> Text
Created
Jun 30, 2026
Reasoning
Publication status
Modalities
Generation controls
Capability scores
Better than 80% of models compared
Better than 77% of models compared
Better than 80% of models compared
Agent tasks
Category performance
1238 Elo
54.4% win
1225 Elo
54.4% win
1249 Elo
53.7% win
1217 Elo
53.7% win
1235 Elo
53.3% win
1226 Elo
52.3% win
1210 Elo
50.9% win
1216 Elo
50.1% win
Creative tasks
Category performance
1263 Elo
54.0% win
1279 Elo
53.8% win
1252 Elo
52.9% win
1277 Elo
52.6% win
1275 Elo
52.3% win
1281 Elo
51.9% win
1209 Elo
51.5% win
1192 Elo
49.7% win