Kimi K3 vs Fable Coding Comparison
ssh4net · x · 2026-07-19
Benchmarked on a real bug in Cline using the same harness: both **Kimi K3** and **Fable 5** can fix the issue, but with different strengths. - **Fable 5 is faster**: total time **3.5 minutes**, **18** tool calls; Kimi K3 took **12 minutes**, **34** tool calls. - **Kimi K3 is cheaper**: used **1.2M** tokens, costing **$0.92**; Fable 5 used **730K** tokens, costing **$2.13**. - The author believes the traces show Kimi is more like a model trained with reinforcement learning, spending more tokens on thinking and verification. - Conclusion: this is the first time they've seen an **open-weight model** genuinely compete head-to-head with SOTA in a real coding-agent scenario.
Related event: Kimi K3 Evaluations Show Polarized Results and Harness Sensitivity(5 posts)→
More from coding & agent
- uv-scripts/ocr returns to the top of Hugging Face datasets with a JSON model picker — vanstriendaniel · 2026-07-21
- Sonar CEO says a guide-verify-solve loop cuts coding-agent issues by 92% — alex_verem · 2026-07-21
- OpenAI’s Codex app is good enough that one user would pay for it even with open models — drdanielbender · 2026-07-21
- A creator built an Awwwards-style landing page with ChatGPT 5.6 Sol in one conversation — paw_lean · 2026-07-21
- A creator rebuilt a personal website as a Jupyter Notebook with Codex — paw_lean · 2026-07-21
- OpenAI’s Build Week buildathon drew 40 people for 11 hours with Codex — paw_lean · 2026-07-21