Model comparison
GPT-6 Sol vs GPT-5.6 Terra: same input price, cheaper output
GPT-6 Sol matches GPT-5.6 Terra's input and cache prices and charges less for output. What that means for coding sessions and the move Codex suggests.
· Prices as of September 28, 2026
GPT-6 Sol
OpenAI · Released September 22, 2026
The mid-priced GPT-6 model, which OpenAI pitches for complex coding and agent workflows and which the Codex docs recommend for complex coding.
GPT-6 Sol facts and comparisonsGPT-5.6 Terra
OpenAI · Released July 9, 2026 · Previous generation
The middle tier of GPT-5.6, balancing intelligence and cost, which roughly corresponds to earlier GPT-5 mini models.
GPT-5.6 Terra facts and comparisons
The short answer
GPT-6 Sol costs the same as GPT-5.6 Terra for input, cache writes, and cache hits, and 17% less for output, so the example agentic coding session costs $2.10 against $2.20. Codex suggests moving from GPT-5.6 Terra to GPT-6 Sol, which the Codex docs recommend for complex coding. Terra stays in the API, so the main reason to stay is a pipeline you have already tuned to it.
Choose GPT-6 Sol if
- You want the Codex docs' pick for complex coding at no higher rate than Terra on any line.
- Your work is output-heavy: the example generation costs $0.86 on GPT-6 Sol against $1.02 on Terra.
- You want the newer generation, released on September 22, 2026, which OpenAI pitches for complex coding and agentic workflows.
Choose GPT-5.6 Terra if
- You have a tuned workflow on GPT-5.6 Terra, and it stays available in the API.
- You pick models in Cursor, whose models and pricing page lists GPT-5.6 Terra, and you have not confirmed GPT-6 Sol there.
- You replaced an earlier GPT-5 mini model with Terra and want to keep that GPT-5.6 slot for now.
Side by side
Specs and prices
| Fact | GPT-6 Sol | GPT-5.6 Terra |
|---|---|---|
| Maker | OpenAI | OpenAI |
| API model id | gpt-6-sol | gpt-5.6-terra |
| Released | September 22, 2026 | July 9, 2026 |
| Status | Current | Previous generation |
| Context window | 1.05M tokens | 1.05M tokens |
| Max output | 128K tokens | 128K tokens |
| Open weights | No | No |
| Input, per 1M tokens | $2 | $2 |
| Cache hit, per 1M | $0.20 | $0.20 |
| Cache write, per 1M | $2.50 | $2.50 |
| Output, per 1M tokens | $10 | $12 |
| Runs in | Codex, OpenCode, OpenRouter, and GitHub Copilot | Codex, Cursor, OpenCode, OpenRouter, and GitHub Copilot |
Standard API rates in US dollars, as published by each maker on September 28, 2026. Batch and priority tiers, taxes, and subscription plans are not included. GPT-6 Sol: Requests over 272K input tokens cost 2x for input and cache and 1.5x for output, for the whole request. GPT-5.6 Terra: Requests over 272K input tokens cost 2x for input and cache and 1.5x for output, for the whole request.
Cost
What the same work costs
The same token counts, priced at each model’s published rates. Cached tokens are billed at each maker’s cache prices, so the session shows what caching is worth on each model.
Real sessions differ: the two models count the same code as different numbers of tokens, and reasoning settings change how much each one writes. Your own history is the real test.
| Workload | GPT-6 Sol | GPT-5.6 Terra |
|---|---|---|
| Agentic coding session, 100K input, 400K written to cache, 2M read from cache, 50K output | $2.10 | $2.20 |
| Large one-off review, 150K input with no cache hits, 10K output | $0.40 | $0.42 |
| Output-heavy generation, 30K input, 80K output | $0.86 | $1.02 |
| A month of sessions, 110 sessions: 5 a day, 22 working days | $231.00 | $242.00 |
| Where the session’s cost goes | ||
| Cache writes | $1.00 | $1.00 |
| Cache reads | $0.40 | $0.40 |
| Uncached input | $0.20 | $0.20 |
| Output | $0.50 | $0.60 |
- caching saves on the session with GPT-6 Sol (62%)
- $3.40
- caching saves on the session with GPT-5.6 Terra (61%)
- $3.40
Why GPT-6 Sol and GPT-5.6 Terra cost almost the same
GPT-6 Sol and GPT-5.6 Terra share three of their four rates: $2 per million input tokens, $0.20 per million cached tokens, and $2.50 per million cache-write tokens. Output is the one difference, $10 per million on GPT-6 Sol against $12 on Terra.
Because only output differs, the gap tracks how much a job writes. The session costs $2.10 against $2.20, a 5% gap worth $0.10. The uncached review costs $0.40 against $0.42. The output-heavy generation, where 80K tokens come back, costs $0.86 against $1.02, 16% less on GPT-6 Sol.
Everything else in the session matches to the cent. Cache writes cost $1.00 on each, cache reads $0.40, and fresh input $0.20, and caching saves $3.40 on each. Over 110 sessions a month, GPT-6 Sol comes to $231.00 and Terra to $242.00, $11.00 apart.
A newer tier at the old middle tier's price
The two sit in different places in their families. Terra was the middle of three GPT-5.6 tiers, which OpenAI described as balancing intelligence and cost and placed roughly where earlier GPT-5 mini models sat. GPT-6 Sol is the mid-priced GPT-6 model, which OpenAI says is built to power complex coding and agentic workflows.
So the upgrade Codex suggests keeps the price and moves up OpenAI's positioning. OpenAI claims GPT-6 Sol makes about half as many mistakes as its predecessor on its internal factuality evaluation, and the Codex docs recommend it for complex coding. These are the maker's claims, so run your own tasks before moving a team.
Terra reached its current price through a 20% cut on July 30, 2026. Neither model's rates carry a promotional note in OpenAI's pricing.
What changes when you switch from GPT-5.6 Terra to GPT-6 Sol
Caching rules stay the same, since both come from GPT-5.6 or later: writes at 1.25x input, hits at 0.1x, up to four explicit cache breakpoints, and prefixes that stay reusable for at least 30 minutes after their last use. Limits match as well, with a 1.05M context window and 128K of output on each, and the same premium above 272K input tokens: 2x for input and cache, 1.5x for output. GPT-6 Sol accepts up to 922K input tokens of its window.
Reasoning effort is the setting to check. GPT-6 Sol defaults to medium effort in the API and in Codex. The sources here do not give Terra's default, so compare your settings after switching, because effort changes how many output tokens each request writes.
Both run in Codex, OpenCode, OpenRouter, and GitHub Copilot. Cursor's models and pricing page lists Terra, and the sources here don't say whether Cursor offers GPT-6 Sol.
To compare them on your own work rather than on example workloads, EveryToken prices your local Codex and OpenCode requests at API rates and splits them by model.
Prompt caching
How OpenAI bills cached tokens
OpenAI
Prompt caching is on by default. From GPT-5.6 on you can also mark up to four explicit cache breakpoints, while GPT-5.5 and earlier cache automatically only.
From GPT-5.6 on, a cache write costs 1.25x the uncached input price and a cache hit costs 0.1x. GPT-5.5 and earlier add no charge for writing the cache: written tokens are billed as ordinary input, and a hit costs 0.1x on the models compared here.
On GPT-5.6 and later, a cached prefix stays reusable for at least 30 minutes after its last use, and caching starts at 1,024 input tokens.
Source: OpenAI: Prompt caching
Your own numbers
See what GPT-6 Sol and GPT-5.6 Terra really cost you.
everyaitoken reads your Codex, OpenCode, OpenRouter, and Cursor history on your Mac and prices every request at API rates, with what caching saved or cost. $9 once.
FAQ
Questions
Is GPT-6 Sol more expensive than GPT-5.6 Terra?
No. It costs the same for input, cache writes, and cache hits, and less for output: $10 per million tokens against $12. In the example session that works out to $2.10 against $2.20.
Does Codex recommend moving from GPT-5.6 Terra to GPT-6 Sol?
Yes. Codex suggests moving from GPT-5.6 Terra, as well as GPT-5.6 Sol and GPT-5.4, to GPT-6 Sol. Terra stays available in the API.
Will my cache costs change if I switch?
Not per token. Both bill cache writes at $2.50 and cache hits at $0.20 per million tokens. In the example session, cache writes cost $1.00 and cache reads $0.40 on both.
Is GPT-5.6 Terra being retired?
Not from the API, where it remains as a previous-generation model, though Codex suggests GPT-6 Sol instead.
Sources
- OpenAI: API pricing
- OpenAI docs: GPT-6 Sol
- OpenAI: Introducing GPT-6 Sol and Luna
- OpenAI: API changelog
- Codex docs: Models
- OpenRouter: GPT-6 Sol
- OpenCode docs: Zen
- GitHub Docs: Supported AI models in Copilot
- OpenAI docs: GPT-5.6 Terra
- OpenAI: Using GPT-5.6
- Cursor docs: Models and pricing
- OpenRouter: GPT-5.6 Terra
- OpenAI: Prompt caching