We’re excited to announce Composer 2, our most capable coding model yet!
Composer 2 delivers frontier-level performance on coding tasks with substantial improvements across all benchmarks. It’s priced at $0.50/M input and $2.50/M output tokens. There is also a faster variant with the same intelligence at $1.50/M input and $7.50/M output.
Key improvements:
- Frontier-level coding intelligence: Composer 2 shows large improvements on all benchmarks we measure, including Terminal-Bench 2.0 and SWE-bench Multilingual.
- Long-horizon reasoning: Built on our first continued pretraining run, Composer 2 is trained on long-horizon coding tasks through reinforcement learning. It can solve challenging problems requiring hundreds of actions.
- Two speed options: We’re introducing a faster variant with the same intelligence. Fast is the default option and has a lower cost than other fast models.
It’s available now!
We’d also love for you to share your first impressions here!