AIster · The model wars · Launches
Anthropic releases Claude 3.5 Sonnet
A mid-priced model outperforms Anthropic's own flagship and arrives with Artifacts
Key facts
- Date
- 20 Jun 2024
- Era
- The model wars
- Interest
- Launches
- Threads
- Model wars
- Who
- Anthropic
- Price
- $3 / $15 per million input / output tokens; 200K context
- Benchmark
- Internal agentic coding eval: 64% vs Claude 3 Opus 38%
Anthropic's Claude 3.5 Sonnet outperformed its own flagship, Claude 3 Opus, at twice the speed and the mid-tier price of $3 per million input tokens (the chunks of text a model reads). In an internal agentic coding test, where the model works through tasks on its own, it solved 64% of problems against 38% for Opus. It launched alongside Artifacts, a panel for editing code and documents Claude creates.
Sources
- Anthropic, Introducing Claude 3.5 Sonnet, 2024-06-21
- TechCrunch, Anthropic claims its latest model is best-in-class, 2024-06-20