What is GPT-5.6 Luna?

GPT-5.6 Luna is OpenAI's budget-friendly model in the GPT-5.6 family (Sol/Terra/Luna), made generally available July 9, 2026. It's optimized for speed and general tasks, and became the free ChatGPT default after an ~80% price cut on July 30, 2026, making it the cheap-tier workhorse for everyday queries.

Which cheap AI model is best in August 2026?

GPT-5.6 Luna is the best free ChatGPT default for general use. Gemini 3.6 Flash is Google's most efficient model — best for Google-native workflows and multimodal. DeepSeek V4 Flash is the cheapest via API and best for developers wanting rock-bottom token costs. Pick Luna for free chat, Gemini 3.6 Flash for Workspace, DeepSeek V4 Flash for cheapest API.

What is Gemini 3.6 Flash?

Gemini 3.6 Flash is Google's most efficient model, made generally available July 21, 2026 alongside Gemini 3.5 Flash-Lite. It's designed to streamline reasoning steps and tool calls, making it fast and cheap for high-volume, everyday tasks with strong multimodal support.

Quick Answer

GPT-5.6 Luna vs Gemini 3.6 Flash vs DeepSeek V4 Flash

Published: August 10, 2026

The Short Answer

As of August 2026, the top budget-tier models are GPT-5.6 Luna (free ChatGPT default), Gemini 3.6 Flash (Google’s most efficient), and DeepSeek V4 Flash (cheapest API). Luna wins for free chat, Gemini 3.6 Flash for Workspace/multimodal, DeepSeek V4 Flash for lowest API cost.

Quick Comparison

	GPT-5.6 Luna	Gemini 3.6 Flash	DeepSeek V4 Flash
Vendor	OpenAI	Google	DeepSeek
GA date	Jul 9, 2026	Jul 21, 2026	2026
Free access	ChatGPT free default	Gemini free tier	API only
Core strength	Speed, general chat	Efficiency, multimodal	Rock-bottom API cost
Best for	Everyday free chat	Google-native tasks	Cheapest developer API

GPT-5.6 Luna — Free ChatGPT Default

Luna is the budget tier of the GPT-5.6 family (Sol/Terra/Luna), GA July 9, 2026. After an ~80% price cut on July 30, 2026, it became the free ChatGPT default — optimized for speed and general tasks. It’s the pick for everyday free chat and high-volume light queries.

Gemini 3.6 Flash — Most Efficient

Gemini 3.6 Flash (GA July 21, 2026) is Google’s most efficient model, designed to streamline reasoning steps and tool calls. Fast and cheap for high-volume everyday tasks with strong multimodal support. Best for teams inside Google Workspace and multimodal pipelines.

DeepSeek V4 Flash — Cheapest API

DeepSeek V4 Flash targets developers who want rock-bottom token costs via API. It’s the cheapest of the three for programmatic use, and pairs with DeepSeek V4’s open-weight lineage for self-hosting flexibility. Best for cost-sensitive developer workloads at scale.

Which Should You Pick?

Free everyday chat → GPT-5.6 Luna.
Google Workspace + multimodal → Gemini 3.6 Flash.
Cheapest API for developers → DeepSeek V4 Flash.

The Reality Check

All three are budget tiers — for frontier reasoning, coding, or long-horizon agentic work, step up to GPT-5.6 Sol, Claude Opus 5, or Gemini 3.5 Pro (when GA). The cheap tier is for volume and everyday tasks, not hardest-problem reasoning.

Sources

OpenAI — GPT-5.6 family (Jul 9, 2026): openai.com
Google — Gemini API changelog (3.6 Flash GA): ai.google.dev
DeepSeek — API pricing: api-docs.deepseek.com