GPT-5.6 API Status: OpenAI GA, Model IDs, and EvoLink Route Checklist
gpt-5.6, gpt-5.6-sol, gpt-5.6-terra, and gpt-5.6-luna.Quick Status
| Question | Current status | What it means for you |
|---|---|---|
| Is GPT-5.6 official? | Yes. OpenAI GA was announced on July 9, 2026. | You can plan production adoption instead of preview-only experiments. |
| Is it one model? | No. GPT-5.6 is a family: Sol, Terra, and Luna. | Route by workload instead of hard-coding one tier everywhere. |
What does gpt-5.6 mean? | OpenAI lists it as an alias to gpt-5.6-sol. | Use the alias only when Sol is intentionally the default. |
| Is GPT-5.6 live on EvoLink? | Yes. The GPT-5.6 route is live. | Confirm pricing and account access in your EvoLink dashboard before scaling traffic. |
| Which tier should I pick? | Depends on your workload. | Use the Sol vs Terra vs Luna routing guide for model choice. |
Confirmed Model IDs
| OpenAI model ID | Tier | Use when |
|---|---|---|
gpt-5.6 | Alias to Sol | You want the GPT-5.6 flagship behavior and accept Sol-level cost. |
gpt-5.6-sol | Flagship | Frontier reasoning, hard coding tasks, deep research, long-horizon agents. |
gpt-5.6-terra | Balanced | Everyday production traffic where quality, latency, and cost all matter. |
gpt-5.6-luna | Cost-efficient | High-volume, low-latency, budget-controlled workloads. |
Pricing Boundary
OpenAI list pricing is:
| Tier | Input / 1M tokens | Output / 1M tokens | Production note |
|---|---|---|---|
| Sol | $5.00 | $30.00 | Use for high-value reasoning and agentic tasks. |
| Terra | $2.50 | $15.00 | Test as the default production candidate. |
| Luna | $1.00 | $6.00 | Use for high-volume tasks when quality is sufficient. |
Do not copy these values into customer-facing EvoLink quotes without checking EvoLink route pricing, account terms, cache discounts, retries, and actual successful-task cost.
What Still Needs Verification Before Production
| Area | Why it matters |
|---|---|
| EvoLink route pricing | OpenAI list price is not always the same as platform route pricing. |
| Account access | Availability can vary by account, region, or provider-side permissions. |
| Cache billing | GPT-5.6 workloads may depend heavily on cached input economics. |
| Rate limits | Sol, Terra, and Luna can need different rollout limits. |
| Fallback behavior | Production traffic should not depend on one new route without a rollback path. |
| Request surface | Confirm supported endpoints, streaming, tools, and reasoning controls against EvoLink docs before publishing code examples. |
Recommended Next Step
Official Sources
Related Reading
FAQ
Is GPT-5.6 generally available?
Yes. OpenAI announced GPT-5.6 general availability on July 9, 2026.
What is the GPT-5.6 model ID?
gpt-5.6 as an alias to gpt-5.6-sol. The tier-specific model IDs are gpt-5.6-sol, gpt-5.6-terra, and gpt-5.6-luna.Is GPT-5.6 available on EvoLink?
EvoLink now positions GPT-5.6 as a live model route. Before production rollout, verify account access, pricing, limits, and fallback behavior inside EvoLink.
How much does GPT-5.6 cost?
OpenAI list pricing is $5/$30 for Sol, $2.50/$15 for Terra, and $1/$6 for Luna per 1M input/output tokens. EvoLink route pricing should be confirmed before quoting customers.
Should teams migrate from GPT-5.5 immediately?
Not blindly. Keep GPT-5.5 as a baseline and fallback while testing GPT-5.6 Sol, Terra, and Luna on real workloads.
Which GPT-5.6 tier should be default?
Terra is the most natural default candidate for balanced production traffic, but each team should benchmark Sol, Terra, and Luna against task value, latency, and cost.
Should code examples hard-code gpt-5.6-sol?
Only when Sol is intentionally required. Otherwise keep model choice configurable so teams can route between Sol, Terra, Luna, and fallbacks without code changes.
