GPT-5.6
OpenAI's GPT-5.6 model family (Sol, Terra, Luna), since followed by GPT-6
OpenAI has replaced this model with GPT-6.1 Sol.
Context window
1.05M
Input / 1M tokens
$4.00
Output / 1M tokens
$20.00
Provider
OpenAI
Pricing is set per tier, per 1M input/output tokens: GPT-5.6 Sol $4/$20 under promotional pricing that OpenAI says runs until November 30, 2026 (launch price $5/$30), GPT-5.6 Terra $2/$12 and GPT-5.6 Luna $0.20/$1.20. Prompts over 272K input tokens cost 2x input and 1.5x output. Cache writes are billed at 1.25x the uncached input rate. GPT-6.1 Sol ($2/$10), GPT-6.1 Terra ($1/$5) and GPT-6 Luna ($0.10/$0.50) cost less. · Data verified 2026-11-08
GPT-5.6 is the OpenAI model family previewed on June 26, 2026 that introduced the current naming system: the number marks the generation, while Sol, Terra and Luna mark capability tiers spanning intelligence, speed and cost. Sol is the flagship (the gpt-5.6 alias routes to gpt-5.6-sol), Terra is the mid tier and Luna the most economical. The family advanced software engineering, computer use, knowledge work and cybersecurity, and added more predictable prompt caching. After a government-coordinated limited preview it became generally available in the API. OpenAI has since released the GPT-6 family (GPT-6 Astra, GPT-6.1 Sol, GPT-6.1 Terra and GPT-6 Luna), which costs less per token; the GPT-5.6 models remain in the API with no announced shutdown date.
Capability index
Relative estimates (0-100) to place this model against its peers, grounded in published benchmarks.
How to access it
Generally available in the OpenAI API by tier: gpt-5.6-sol (also reachable through the gpt-5.6 alias), gpt-5.6-terra and gpt-5.6-luna, on the Chat Completions, Responses and Batch endpoints; GPT-5.6 Sol is also on Amazon Bedrock (openai.gpt-5.6-sol). ChatGPT has moved on to GPT-6 models.
Strengths
- ✓Tiered lineup (Sol/Terra/Luna) lets teams trade off intelligence, speed, and cost within one generation
- ✓Frontier gains in software engineering, computer use, and cybersecurity
- ✓More predictable prompt caching (explicit breakpoints, 30-minute minimum cache life)
- ✓Improved token efficiency over the GPT-5.5 generation
Best for developers who...
When to choose it (and when not to)
Reach for GPT-5.6 when...
- →You have workloads tuned to a GPT-5.6 tier and want stable behavior
- →You want to route requests across Sol, Terra and Luna within one generation
Look elsewhere if...
- ✕New projects: the GPT-6 tiers cost less per token (GPT-6.1 Sol $2/$10, GPT-6 Luna $0.10/$0.50)
- ✕You need a stable price: GPT-5.6 Sol's $4/$20 is promotional pricing that OpenAI says runs until November 30, 2026
- ✕You need the strongest OpenAI model (use GPT-6 Astra)
How to use it
- ›Pick the tier that matches the task: Sol for hardest reasoning/agentic work, Terra for balance, Luna for high-volume/low-latency
- ›Use explicit cache breakpoints to control caching cost on large repeated prefixes
- ›On Sol, reasoning effort ranges from none to max, with medium as the default
Quickstart
Pythonfrom openai import OpenAI
client = OpenAI()
# Choose a tier explicitly: gpt-5.6-sol / gpt-5.6-terra / gpt-5.6-luna
response = client.responses.create(
model="gpt-5.6-terra",
input="Summarize this contract and flag risky clauses.",
)
print(response.output_text)There is no single 'gpt-5.6' endpoint documented; select a tier (Sol/Terra/Luna). Ids shown are inferred placeholders pending official docs. Access is limited to approved preview partners.
API model id: gpt-5.6-sol
Benchmarks
| Benchmark | Score | Notes |
|---|---|---|
| Terminal-Bench 2.1 (Sol) | 88.8% | Flagship Sol tier of the family. |
| Terminal-Bench 2.1 (Terra) | 82.5% | Mid tier. |
| Terminal-Bench 2.1 (Luna) | 84.3% | Fastest/most economical tier. |
| Terminal-Bench 2.1 | 91.9% | Base Sol; 91.9% with Sol Ultra |
| ExploitBench | Competitive with Anthropic Mythos Preview | Using ~1/3 of output tokens |
| ExploitGym | Significant improvements | Sol, Terra, Luna all improved with reasoning |
| GeneBench v1 | Not announced | Sol achieves stronger results than GPT-5.5 while using fewer tokens |
Compare GPT-5.6
GPT-5.6 vs Claude Sonnet 5
Anthropic - 1M ctx
GPT-5.6 vs Gemini 3.5
Google DeepMind - 1,048,576 tokens (Gemini 3.5 Flash; Pro variant not yet released) ctx
Compare GPT-5.6 with any other model
Build a comparison →All model comparisons →