Blog · 2026-07-10

GPT-5.6 Sol vs Terra vs Luna: Which Tier Should You Actually Use?

OpenAI's GPT-5.6 family ships in three tiers — Sol, Terra and Luna — and they are the same model line at three price points, not three different architectures. All three take text and image input, return text, carry a 1,050,000-token context window with up to 128,000 output tokens, support reasoning tokens, and share a February 2026 knowledge cutoff. Choosing between them is almost entirely a cost decision.

On apimodels.app the three sit at about 8 : 3.4 : 1 on input and 7 : 3.4 : 1 on output. Sol costs $1.324 per million input tokens and $6.618 per million output. Terra is $0.551 and $3.309 — half of Sol's output, about 40% of its input. Luna is $0.16 and $0.96. Cached input tokens are billed at a tenth of the input rate on every tier — $0.132, $0.055 and $0.016 respectively. All three are below OpenAI's own list price; the current per-tier comparison is on /access/gpt-5-6-api-pricing.

A practical rule: start on Terra. It is the tier most workloads should default to — everyday chat, code generation, agentic tool calling — because doubling your spend to reach Sol buys reasoning depth you often will not use. Move up to Sol when a task genuinely fails on Terra: long multi-step planning, hard debugging, research synthesis. Move down to Luna when the job is bulk and mechanical: classification, extraction, rewriting, first-line support. At $0.16 per million input tokens, Luna makes batch pipelines that would be uneconomical on a frontier model suddenly affordable.

Upstream, OpenAI exposes reasoning depth as a suffix on the model name rather than a body parameter — gpt-5.6-sol-low through -ultra are separate model ids, all at the same per-token rate, differing only in how many reasoning tokens they emit. apimodels.app serves the three base tiers only: gpt-5.6-sol, gpt-5.6-terra and gpt-5.6-luna. The depth variants were retired on 2026-07-27 when the channel that carried them shut down, and the replacement upstream does not offer them; calling a suffixed id returns a clear error rather than silently substituting a different model. Pick the tier that matches the job — that is the decision that actually moves cost and quality.

Three behaviours are worth knowing before you build on any of the three. First, the system preamble upstream injects varies by pool — measured 2026-09-05, a one-line prompt reported about 4,400 input tokens (roughly 3,840 of them cache hits, billed at the cached rate), though single-digit requests also occur, so do not budget against a fixed floor even for a one-word prompt — budget for it when you price short, high-frequency calls. Second, parallel_tool_calls: false is ignored: ask about the weather in Paris and Tokyo and you will get two tool_calls back regardless, so write your handler to iterate rather than assume one. Third, once a single request exceeds 272,000 input tokens the entire request is rebilled at 2x input and 1.5x output — the step is on the request, not on the tokens past the threshold.

All three answer on both OpenAI-compatible endpoints, /v1/chat/completions and /v1/responses, with function calling, streaming and multi-turn tool results verified on each. Point base_url at https://api.apimodels.app/v1, keep your existing OpenAI SDK, and switch tiers by editing one string. Two aliases exist for convenience: gpt-5.6 routes to Sol, and gpt-5.6-luna routes to gpt-5.6-luna-max, which is currently the only Luna depth the upstream serves — same price, so nothing changes on your bill.

FAQ