How to Build Tool-Using AI Workflows with the Responses API
A practical guide to custom tools, public MCP servers on supported OpenAI models, streaming tool calls, and stored response chains in NanoGPT's Responses API.
NanoGPT Blog
Updates, guides, and insights
Showing
Qwen3.8 Max Preview: What We Know Before the Benchmarks
Qwen3.8 Max Preview is available before its benchmark table. Here is what is confirmed, what remains unverified, and how to compare it fairly with Qwen3.7 Max.
Doubao Seed Character vs Aion 3: Which Roleplay Model Should You Use?
Compare Doubao Seed Character, Aion 3.0, and Aion 3.0 Mini for character chat, roleplay, long-form storytelling, image input, and price.
Fugu Ultra Benchmarks: When Multi-Agent AI Is Worth the Cost
See how Sakana AI's Fugu Ultra uses multiple AI agents, what its coding and reasoning benchmarks show, what it costs, and when the premium makes sense.
How to Choose an AI Web Search Tool
Compare Linkup, Brave, Tavily, Exa, Kagi, Perplexity, Valyu, Sofya, and Firecrawl by cost, search depth, page content, filters, and best use case.
Muse Spark 1.1 Benchmarks: How Good Is Meta's New Model?
Muse Spark 1.1 brings strong coding results, a 1M-token context window, and inexpensive cached input. Here is where Meta's new model stands out—and where it still falls short.
AI for Roleplay and Long-Form Storytelling with Aion 3.0
Use Aion 3 for immersive roleplay and longer stories with practical prompts for characters, scenes, continuity, pacing, and creative boundaries.
How to Analyze FIT Workout Files with AI
Upload a FIT workout file to NanoGPT and use AI to examine pacing, heart rate, power, laps, elevation, and training patterns responsibly.
GPT-5.6 Sol vs Terra vs Luna: Which One Should You Use?
Compare GPT-5.6 Sol, Terra, and Luna by quality, speed, price, coding performance, and long-context reliability—and see which tier fits your work.
How to Choose the Right Reasoning Effort for AI Models
Learn when to use low, medium, high, or maximum AI reasoning effort—and why more thinking is not always the better choice.