← All IntelIntel / article
Kimi K3 vs Claude Fable 5 on DeepSWE: Cost and Coding
- Source
- www.together.ai
- Date
Why it matters
- A head-to-head on agentic coding that separates raw accuracy from solves-per-dollar, which is the axis that actually decides which model you can afford to run in a loop.
- We ran 452 DeepSWE rollouts on Kimi K3 and Claude Fable 5.
- Fable leads pass@1 by 1.4 points; Kimi K3 wins pass@4 and delivers 2.8x the solves per dollar.
Transcript
We ran 452 DeepSWE rollouts on Kimi K3 and Claude Fable 5. Fable leads pass@1 by 1.4 points; Kimi K3 wins pass@4 and delivers 2.8x the solves per dollar.
Read the source ↗www.together.ai
Recommended reads
Comments
Sign in to comment.
Loading comments…