Official commercial channel
GPT-5.6 Sol API
The frontier tier of the GPT-5.6 family: 1.05M context, built for long-horizon agents.
Context
1.05M
Max output
128K
Serving precision
Vendor-managed
Licence
Proprietary
GPT-5.6 shipped on 9 July 2026 as a family of three — Sol, Terra and Luna — sharing a 1,050,000-token context window, 128,000 maximum output tokens and a 16 February 2026 knowledge cut-off, accepting text and image input. Sol is the frontier flagship, aimed at complex reasoning, coding and long-horizon agentic work, at vendor list pricing of $5 per million input and $30 per million output. We provide it through an official commercial channel — no account pools, no reverse-engineered client protocols.
Proprietary, provided through the vendor’s official commercial licensing. Contract, invoice, audit trail — not an account pool and not a reverse-engineered client protocol.
Specifications
- Model ID
- gpt-5.6-sol
- Published by
- OpenAI
- How it is served
- Official commercial channel
- Licence
- Proprietary
- Context
- 1.05M
- Max output
- 128K
- Serving precision
- Vendor-managed
- Input modalities
- Text · Image
- Released
- 2026-07-09
- Price
- See pricing
What it is good at
- A 1.05M-token window
- A whole codebase, a literature set, months of history in a single request — and all three tiers share the same window, so moving between Sol, Terra and Luna never forces you to redesign your chunking.
- Built for long-horizon agents
- Sol is the tier explicitly aimed at long-horizon agentic work. Multi-step, multi-tool runs that must hold a goal over time are where the price difference against Terra actually earns itself.
- Official channel
- No mass-registered consumer accounts, no forged client identity. The price corresponds to compliant supply, not grey-market quota that can be cut off at any time.
Where it fits
- Complex reasoning and hard coding tasks
- Long-horizon autonomous agent runs
- One-shot processing of very long research reports and literature sets
- Ceiling comparisons against the open-weight flagships
Three steps
Change one line of base_url, put the model ID below in the `model` field, and leave the rest of your code alone.
model: "gpt-5.6-sol"
curl https://api.nexaix.net/v1/chat/completions \
-H "Authorization: Bearer $NEXAIX_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-5.6-sol",
"stream": true,
"messages": [
{ "role": "user", "content": "Summarise this contract." }
]
}'import os
from openai import OpenAI
client = OpenAI(
# the only new line
base_url="https://api.nexaix.net/v1",
api_key=os.environ["NEXAIX_API_KEY"],
)
# everything below is your existing code
stream = client.chat.completions.create(
model="gpt-5.6-sol",
stream=True,
messages=[
{"role": "user", "content": "Summarise this contract."}
],
)
for chunk in stream:
print(chunk.choices[0].delta.content or "", end="")import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api.nexaix.net/v1", // the only new line
apiKey: process.env.NEXAIX_API_KEY,
});
// everything below is your existing code
const stream = await client.chat.completions.create({
model: "gpt-5.6-sol",
stream: true,
messages: [
{ role: "user", content: "Summarise this contract." }
],
});
for await (const chunk of stream) {
process.stdout.write(chunk.choices[0]?.delta?.content ?? "");
}About this model
Sol, Terra or Luna?
All three share the context window and output ceiling; they differ in capability and price. Sol is the frontier tier (list $5/$30 per million tokens), Terra the balanced everyday tier ($2.50/$15), Luna the fastest and cheapest for high volume ($1/$6). The usual pattern is Luna or Terra by default with Sol behind a difficulty check.
Why isn’t your GPT pricing as low as other relays?
Because ours goes through official commercial licensing. GPT relays priced well under official rates are usually supplied by mass-registered consumer accounts or reverse-engineered client protocols — the discount is paid for with sudden outages and with your conversations passing through layers nobody controls.
Does it accept audio or video?
No. All three GPT-5.6 tiers take text and image input and return text; there is no audio or video input. For video understanding use Kimi K3 or MiniMax M3.
Also worth a look
- Official commercial channelClaude Opus 5A million tokens that never degrade, five-level effort control, zero data retention.claude-opus-5
- Official commercial channelClaude Fable 5Anthropic’s tier above Opus — the one you pick when the ceiling is the point.claude-fable-5
- Official commercial channelGPT-5.6 TerraThe everyday tier: the same window as Sol, at half the list price.gpt-5.6-terra
Run your own eval on it
Take the trial credit, run the evals you already have against this model, and then decide whether it gets your production traffic.
