GPT-5.6 Sol, Terra, and Luna Are Now Available on WidelAI

OpenAI's newest model family is live on WidelAI today. GPT-5.6 Sol, GPT-5.6 Terra, and GPT-5.6 Luna are in your model picker right now under the tags gpt-5.6-sol, gpt-5.6-terra, and gpt-5.6-luna — no waitlist, no separate OpenAI subscription, and no API keys to manage. This release replaces the old "unsuffixed / mini / nano" naming with three clearly named tiers that each target a different point on the intelligence-versus-cost curve. This post covers what each model is, how they are priced in WidelAI credits, when to reach for each one, and how to start using them in under a minute.

OpenAI logo

The GPT-5.6 family is built by OpenAI. Read the official model docs for GPT-5.6 Sol, GPT-5.6 Terra, and GPT-5.6 Luna.

Meet the GPT-5.6 Family

GPT-5.6 is OpenAI's latest generation, and it ships as three sibling models that share the same architecture, context window, and knowledge cutoff but differ in reasoning depth, speed, and price. Instead of cryptic suffixes, each tier now has a name:

GPT-5.6 Sol — the frontier model for complex professional work, with the highest reasoning of the family. It roughly corresponds to the unsuffixed model tier used in earlier GPT-5 families, and the gpt-5.6 alias routes to Sol.
GPT-5.6 Terra — the balanced tier that trades a little raw intelligence for meaningfully lower cost. It roughly corresponds to the mini tier of earlier GPT-5 families.
GPT-5.6 Luna — the cost-efficient tier built for high-volume, latency-sensitive workloads. It roughly corresponds to the nano tier of earlier GPT-5 families.

All three share the same core specs:

1,050,000-token context window — over a million tokens, enough for entire codebases, large document sets, or long multi-turn sessions
128,000 max output tokens per request
February 16, 2026 knowledge cutoff
Multimodal input (text and image) with text output
Reasoning token support built in

They also carry the full modern OpenAI tool surface: web search, file search, image generation, code interpreter, hosted shell, apply patch, skills, computer use, MCP, and tool search — plus streaming, function calling, and structured outputs. Fine-tuning is not supported on this family.

How the Three Tiers Compare

The simplest way to think about the family is a single dial from "most capable" to "most economical." All three sit on the same 1M+ context window, so the choice is about reasoning depth and cost, not context size.

Model	Tier	Reasoning	Real API price (input / output per 1M)
GPT-5.6 Sol	Frontier	Highest	$5.00 / $30.00
GPT-5.6 Terra	Balanced	Higher	$2.50 / $15.00
GPT-5.6 Luna	Cost-efficient	High	$1.00 / $6.00

Each model also supports cached input at a steep discount ($0.50, $0.25, and $0.10 per 1M respectively), which helps when you reuse large system prompts or documents across requests. One billing note worth knowing: prompts above 272K input tokens are priced at 2x input and 1.5x output for the full request, so very large single prompts cost more per token than typical usage.

GPT-5.6 Pricing on WidelAI

WidelAI uses a transparent credit system: 1 credit equals $0.005 of real API cost, and input and output tokens are priced separately because generating output costs far more than reading input. Your credit rate for any model is simply its real per-million price multiplied by 0.2, then expressed per 1,000 tokens:

Model	Input (credits / 1K tokens)	Output (credits / 1K tokens)
GPT-5.6 Sol	1.00	6.00
GPT-5.6 Terra	0.50	3.00
GPT-5.6 Luna	0.20	1.20

For reference, GPT-5.6 Sol lands at the same rate as GPT-5.5 (1.00 / 6.00), while Terra matches the older GPT-5.4 pricing and Luna comes in well below both. That means you can move to the newer generation without paying a premium over the models you were already using. See every model side by side on our pricing transparency page.

When to Use Each Model

The three tiers are designed to be mixed and matched. Route each task to the cheapest model that can do the job well, and escalate only when the work demands it.

Reach for GPT-5.6 Sol when

The task is genuinely hard: architecture decisions, tricky debugging, multi-step agentic runs, or deep analysis where getting it right the first time is worth the cost
You are doing large-scale codebase work — migrations and multi-file refactors that need the full million-token context and the highest reasoning
You need long-horizon autonomy without the model drifting or losing state

Reach for GPT-5.6 Terra when

You want strong intelligence for everyday development, writing, and analysis at roughly half the cost of Sol
The workload is substantial but does not require frontier-level reasoning
You are balancing quality against a real budget across many requests

Reach for GPT-5.6 Luna when

Volume and latency matter more than maximum depth: classification, extraction, summarization, routing, and high-throughput chat
You are running many calls and want cost per call to stay low
You need a fast first pass that can escalate to Sol only when confidence is low

Tip: A common pattern is to draft with Luna or Terra, then escalate the hardest cases to Sol. Because all three share one credit balance on WidelAI, you can build that routing logic without juggling separate accounts or keys.

A Cleaner Naming Scheme

If you have been tracking OpenAI's lineup, the Sol / Terra / Luna names are a welcome change. Earlier GPT-5 generations relied on unlabeled base models plus "mini" and "nano" variants, which made it easy to misjudge which model you were actually calling. The GPT-5.6 family maps those same three tiers to memorable names, and the gpt-5.6 alias always points at the flagship (Sol) so your default calls stay on the top model unless you choose otherwise.

As part of this update, WidelAI has also retired a number of older, now-deprecated OpenAI models from the chat model picker, so the models you can select are the ones OpenAI actively recommends. You can still see the full history — including deprecated models, clearly labeled — in our models library.

How to Use GPT-5.6 on WidelAI

Getting started takes about a minute:

Sign in to your WidelAI account, or create one if you are new.
Open the chat and click the model selector.
Choose GPT-5.6 Sol, GPT-5.6 Terra, or GPT-5.6 Luna from the OpenAI group — or paste the model tag (gpt-5.6-sol, gpt-5.6-terra, or gpt-5.6-luna) if you are calling it programmatically.
Send your prompt. Reasoning runs by default, so you get the model's full capability without extra configuration.

The GPT-5.6 family is available on the Pro plan, right next to Claude Fable 5, Claude Opus 4.8, Claude Sonnet 5, and Gemini 3.5 Flash. Browse the full catalog on our models page.

Why Access It Through WidelAI

You could get GPT-5.6 directly from OpenAI — but then you are managing another subscription, another billing relationship, and another set of API keys, all locked to a single provider. WidelAI gives you the entire GPT-5.6 family alongside the best models from Anthropic and Google, under one plan and one credit balance.

That matters for more than convenience. When you can switch between Sol, Terra, Luna, Claude, and Gemini without re-architecting anything, you get to pick the right model for each task on cost and quality — and you stay resilient if any single model changes price or availability. We wrote more about that resilience angle in Claude Fable 5 Is Now Available on WidelAI, after Fable 5 spent a month offline.

Get Started Today

The GPT-5.6 family is live on WidelAI right now. Whether you need Sol's frontier reasoning for a hard problem, Terra's balance for daily work, or Luna's efficiency for high-volume tasks, all three are one click away in your model picker.

Already a member? Open the chat, pick a GPT-5.6 model, and try it on your next task.