Blog
Guide
Best Open Source LLMs in 2026: 6 Worth Deploying in Production
Open-weight LLMs caught up to GPT-5 and Gemini 3 Pro in 2026. We ranked 6 worth running in production — by reasoning, speed, multimodal, and cost-per-token — with the tradeoffs to know before you deploy.
Hendrix Liu · July 29, 2026
Model
Claude 3.5 Haiku vs. Sonnet: speed or power? A comprehensive comparison
A comprehensive comparison of Claude 3.5 Haiku and Claude 3.5 Sonnet covering benchmarks, speed, cost, and when to choose each model for production workloads.
Hendrix Liu · July 29, 2026
Guide
10 Best Helicone Alternatives in 2026 (Post-Mintlify Acquisition)
Helicone alternatives need to replace both the observability layer and the gateway. Compare 10 platforms on what each covers and what it costs.
Hendrix Liu · July 27, 2026
Guide
RAG Evals in Practice: Personal Course Tutor
I built a RAG tutor that answers only from my course PDFs. Here are the per-step evals that made it trustworthy, and what they caught.
Amy Kodama · July 23, 2026Guide
Best Open Source LLMs in 2026: 6 Worth Deploying in Production
Hendrix Liu · July 29, 2026Model
Claude 3.5 Haiku vs. Sonnet: speed or power? A comprehensive comparison
Hendrix Liu · July 29, 2026Guide
10 Best Helicone Alternatives in 2026 (Post-Mintlify Acquisition)
Hendrix Liu · July 27, 2026Guide
RAG Evals in Practice: Personal Course Tutor
Amy Kodama · July 23, 2026Engineering
How to Evaluate AI Agents in Production (Not Just Benchmarks)
Frank Chen · May 21, 2026Engineering
Prompt Versioning Without Evals Is Just Diff Tracking (2026)
Frank Chen · May 19, 2026Product
Single Agent vs Multi-Agent: Why We Rebuilt Our AI Agent
Marcus Huang · May 5, 2026Guide