OpenAI has launched GPT-6 Sol and GPT-6 Luna, 2 new fashions in its GPT-6 household. They sit beneath GPT-6 Astra, which launched earlier this month. OpenAI skilled each with strategies just like Astra’s. The intention is to deliver Astra’s advances to sooner, extra inexpensive fashions.
Deployable at present? Sure. Each fashions are stay within the OpenAI API as gpt-6-sol and gpt-6-luna. They’re API-only fashions, so there are not any weights to self-host.
Three tiers, one recipe
The GPT-6 household now has 3 tiers. Astra is the highest mannequin for the toughest work. Sol targets advanced coding {and professional} duties at decrease value. Luna targets quick, high-volume on a regular basis work.
OpenAI workforce states higher caching and inference let it serve these fashions extra cheaply. It’s reducing Sol and Luna API costs by 50% in opposition to their GPT-5.6 promotional pricing.
| Mannequin | Enter (per 1M tokens) | Output (per 1M tokens) |
|---|---|---|
| GPT-6 Astra | $10.00 | $50.00 |
| GPT-6 Sol | $2.00 (was $4) | $10.00 (was $20) |
| GPT-6 Luna | $0.10 (was $0.20) | $0.50 (was $1.20) |
One element is price noting. Luna’s output value falls from $1.20 to $0.50, a reduce of about 58%, not 50%.
Benchmarks: what OpenAI reviews
Skilled work: On AutomationBench 1.0.6, Sol at xhigh effort scores 33.2% at $0.27 per activity. Claude Opus 5 at max effort scores 26.9% at 11.1x that value. Low-effort Astra scores 30.3% at 3.9x Sol’s value. Luna at excessive effort beneficial properties 5.4 factors over its predecessor at 58% decrease value per activity.
On Brokers’ Final Examination, Sol at max effort scores 56.4%. That beats Claude Opus 5’s finest rating at 60% decrease value per activity.
Coding: On DeepSWE v1.1, Sol at max effort scores 68.8%. That’s 1.1 factors behind Claude Fable 5 at xhigh, at about 80% decrease value per activity. Luna at max effort scores 66.6%, corresponding to Opus 5 and Fable 5 at medium effort. In these comparisons, Luna prices 93% much less per activity than Opus 5 and 96% lower than Fable 5.
On FrontierCode 1.1 Predominant, which grades whether or not code is able to merge, Sol matches Claude Fable 5.1 at xhigh at a lot decrease value.
Pc use: On OSWorld 2.0 offline, Sol at xhigh scores 60.5% versus 60.3% for Opus 5 at medium. Sol’s value per activity is about 80% decrease. Luna at max beats GPT-5.6 Sol at medium for 1/10 of the price.
Factuality: OpenAI’s inside take a look at makes use of de-identified ChatGPT conversations the place customers flagged mannequin errors. Sol makes about half as many errors as its predecessor. Luna at greater effort matches GPT-5.6 Sol at about 1/100 of its value.
OpenAI additionally carried Astra’s communication type over. Anticipate clearer, barely shorter solutions with much less jargon, particularly in coding conversations.
Immediate caching for long-running brokers
Brokers resend the identical directions, instruments and historical past on each flip. GPT-6 ships an improved immediate caching system with greater cache hit charges by default. Cached enter reads get reductions of as much as 90%. Eligible shared prefixes reused inside a 30-minute window now qualify.
New controls for builders:
- A Immediate Caching Dashboard tracks hit charges over time.
- A diagnostics instrument explains misses, for instance
"purpose": "tools_changed". - Specific breakpoints allow you to select the place a cached prefix ends.
- Reasoning effort can change mid-conversation through
configuration_updatewith out breaking cache. allowed_toolsrestricts callable instruments whereas conserving definitions steady.- Prewarming prepares identified context earlier than the primary consumer request.
The total immediate caching information covers every sample. GitHub reviews these modifications reduce the share of immediate tokens needing recent processing by greater than 50%, serving to Copilot reply sooner.
Availability
- API:
gpt-6-solandgpt-6-luna. - ChatGPT Work and Codex: Plus, Professional, Enterprise, Enterprise and Edu customers.
- Free and Go: Luna within the ChatGPT desktop app.
- Not but in Chat. The ChatGPT rollout is gradual by launch day.
