01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A
STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK
NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX
ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW
GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5
SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01
KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE
AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA
COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT
SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200
SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL
0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI
DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI
FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS
GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE
OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN
01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A
STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK
NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX
ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW
GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5
SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01
KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE
AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA
COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT
SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200 SATELLITE GW ORBIT COLOSSUS FLUX NEBULA AI DOWNLINK STARGATE KIMI 0x4CFF7A 01 SIGNAL SCAN OPUS-5 GB200
SIGNAL AI
RegulusRasalas AdhaferaAlgieba ZosmaDenebola Chertan
Leo
Constellation
ORBITALDownlink
FEATURED

OpenAI's GPT-6.2 Astra and Sol: cheaper to run, top marks on OpenAI's tests, and a rough first day with early testers. What changed and what it means for your business

Updated 18 Nov 2026, 17:02 SAST

The short version: OpenAI released two new artificial intelligence (AI) models on Tuesday, 17 November: GPT-6.2 Astra, its new top model for complex work, and GPT-6.2 Sol, a cheaper model for everyday tasks[1]. Both are available in ChatGPT Work and Codex for paying users, and to developers through OpenAI's application programming interface (API), the service that lets other software use the models[1]. OpenAI cut Astra's API input and output prices by 20% compared with GPT-6 Astra, and kept Sol at its predecessor's prices[1]. OpenAI says Astra sets new highs on several professional and scientific tests[1], and independent testers cited by The Decoder broadly back its results in science and its value for money[5]. But several well-known early testers were scathing. VentureBeat reported that investor Matt Shumer called the models "essentially unusable"[4], and that developer Theo Browne called Astra "the worst model OpenAI has ever released"[4]. Android Authority reported that an OpenAI executive said some reviewers had missed changes made in the final days before launch[6]. Our analysis looks at what this means for South African businesses choosing AI tools.

What OpenAI released

OpenAI describes GPT-6.2 Astra as its "new frontier model for your hardest projects", and GPT-6.2 Sol as its "efficient, versatile model for everyday work"[1]. The company says both models "deliver better results on demanding tasks with fewer output tokens"[1]. A token is a small chunk of text, roughly a short word or part of a word, and it is the unit that AI companies use to measure and charge for usage. Fewer tokens for the same task means lower cost and, on subscription plans, more work within the same usage limits.

Astra succeeds GPT-6 Astra, which OpenAI introduced in September, according to Android Authority[6], and Sol succeeds GPT-6.1 Sol[1]. The new versions came out eight days after Anthropic launched Claude Fable 5.2, according to VentureBeat[4]. We covered the previous fortnight's releases in our OpenAI round-up and the Fable 5.2 launch in our Claude round-up.

GPT-6.2 Astra: the flagship

OpenAI positions Astra for complex professional work. It says Astra's slides, spreadsheets and documents "read less like AI" and follow the templates and conventions of fields such as law, sales, finance and accounting more closely[1]. The company also says Astra writes cleaner code and builds better-looking websites, games and 3D designs[1].

OpenAI's own help page gives practical examples: tracing a bug across several connected systems and fixing it, building a slide deck from a report in a company's own template, and comparing research sources to find where the evidence disagrees[2].

GPT-6.2 Sol: the everyday model

Sol is the cheaper model. OpenAI says it is "smarter, more efficient" than GPT-6.1 Sol, and that it "lets you complete more tasks on the same paid plan"[1]. The company says many of its own staff use Sol as their default model for technical and non-technical work[1]. OpenAI's suggested uses for Sol include reviewing code changes, revising code after feedback, and turning meeting notes into a project brief[2].

According to OpenAI's developer documentation, Sol accepts text and images, produces text, and has a context window of 1,050,000 tokens[3]. The context window is how much material the model can consider at once. Its knowledge cut-off date is listed as 1 June 2026[3], which means it will not know about events after that date unless it is given the information or can search the web.

Where you can use them

OpenAI's announcement says both models are available "to all Plus, Pro, Enterprise, and Edu users in ChatGPT Work and Codex"[1]. ChatGPT Work is the part of ChatGPT built for longer, multi-step tasks, and Codex is OpenAI's coding agent. OpenAI's help page also lists the Business plan for ChatGPT Work[2]. For company accounts, the help page says Enterprise and Edu administrators must switch the models on in workspace settings first[2].

The help page adds a detail that matters for everyday users. On supported consumer plans, new Work chats start with Sol at a "Light" reasoning setting, while Enterprise workspaces with access start with Sol at "Medium"[2]. In other words, unless you change the setting, you may be using the cheaper model with less thinking time, not Astra. To use either model, users open Work, not the standard chat screen[2].

OpenAI is also offering a faster mode it calls Ultrafast. Subscribers on the $500 Pro plan can use it with both models, and for the first time subscribers on the $200 Pro plan can use it with Sol[1][6].

Prices: what it costs to run

For businesses and developers who pay per use through the API, OpenAI lists these standard rates per million tokens[1]:

Model Input Cached input Output
GPT-6.2 Astra $8 $0.40 $40
GPT-6.2 Sol $2 $0.10 $10

"Input" is what you send to the model, "output" is what it writes back, and "cached input" is repeated input that the system has stored, such as the same long instructions sent with every request. OpenAI says Astra's input and output prices are 20% lower than GPT-6 Astra's, with a 60% discount on cached input[1]. Sol keeps GPT-6.1 Sol's prices[1].

VentureBeat's analysis adds several details from OpenAI's pricing documentation[4]:

  • Long requests cost more. For requests with more than 272,000 input tokens, OpenAI doubles the input and cache rates and charges 1.5 times the output rate for the whole request. VentureBeat calculates that this puts a long Astra request at $16 per million input tokens and $60 per million output tokens.
  • Batch is cheaper, speed is dearer. VentureBeat reports half-price batch processing, double the standard rate for the "Fast" tier, and an Ultrafast API tier that starts at three times the standard price and rises to six times on 1 March 2027.
  • Matching rivals. VentureBeat reports that Astra's standard rates are the same as those of Anthropic's Claude Fable 5.2, at $8 and $40 per million tokens, and that Sol's are the same as Anthropic's Sonnet[4].

OpenAI also recommends that developers keep the start of their prompts unchanged and add new instructions at the end, which it says improves cache reuse and lowers cost[1].

What OpenAI claims the models can do

All the figures in this section come from OpenAI's announcement and are the company's own results, unless we say otherwise. Independent checks are covered in the next section.

Professional work

On a test called Agents' Last Exam, which OpenAI says measures multi-step professional work, the company says Astra scored 5.2 points higher than GPT-6 Astra "at less than a third of the cost", and 5.5 points ahead of Claude Fable 5.2 "at just 5% of the cost"[1]. On its internal GDPval v2 test of long professional projects in fields such as finance, data science and marketing, OpenAI says Astra gained 6.2 percentage points over GPT-6 Astra while using 22% fewer output tokens[1].

On spreadsheets, OpenAI cites an Artificial Analysis test using real reports, on which it says Astra answered 62.5% of questions correctly on all five independent attempts[1]. On slides, OpenAI says Astra can match a template's layout, colours and typography "down to the exact font size"[1].

On Harvey LAB-AA, a test of legal tasks such as reviewing contract clauses, OpenAI says Astra more than doubled Claude Fable 5.2's pass rate at about one-sixth of the cost[1]. VentureBeat notes that OpenAI's own chart puts Astra's best pass rate at about 15%, because the test only counts a pass when every criterion is met with no material errors[4]. That is a reminder that "best ever" can still mean most attempts fail on a very strict test.

OpenAI says it believes Astra is "the best model in the world for scientists", and says it scored 9 percentage points higher than GPT-6 Astra on Terminal-Bench-Science, a test of scientific work done at a computer's command line[1]. On a 3D design test run by Artificial Analysis, where human judges compare two blind results, OpenAI reports Astra ranked first, ahead of Claude Fable 5.2 and GPT-6.2 Sol[1].

Statements from customers

OpenAI's announcement includes statements from companies that tested the models, including Harvey, Legora, Ramp, IMC and Jump Trading[1]. These are early-access customers quoted by OpenAI, so they should be read as supportive testimony, not independent reviews. One example: OpenAI quotes Higgsfield AI's chief executive, Alex Mashrabov, saying that in Higgsfield's internal 3D tests Astra reached the target result in one attempt in 73% of cases[1].

What independent testers found

Strong on science and value

The Decoder, an AI news site, summarised results from several independent test groups[5]. It reports that:

  • Artificial Analysis ranks Astra second on its Intelligence Index, just behind Claude Fable 5.2, and says OpenAI's models lead on cost against performance. It also says Sol, at 57 points, matches GPT-6 Astra's level at about 15% of the cost per task[5].
  • Epoch AI, a research organisation, gave Astra a score of 174 on its capabilities index, tying Fable 5.2[5].
  • Vals AI ranks Astra seventh and Sol ninth on its Vals Index, and calls Sol the cheapest model in its top 10 at $3.18 per task, against $10.68 for Astra[5].

Flat on coding, weaker in places

The same independent tests found weaknesses. The Decoder reports that, according to Artificial Analysis, Astra's coding-agent score was unchanged from GPT-6 Astra and Sol's fell by three points from GPT-6.1 Sol, mainly because Sol more often left out key facts when answering questions about code[5]. VentureBeat reports the same finding[4].

The Decoder also reports that Epoch AI found Astra fell behind Fable 5.2 on real research tasks and "barely improved" on GPT-6 Astra there[5]. On legal research, it reports that Vals AI ranked Astra 30th and Sol 52nd, although OpenAI's specialised "Astra for Law" version did better[5].

The Decoder further reports that Artificial Analysis measured high rates of guessing on very hard factual questions: 59% for Astra and 55% for Sol at maximum effort[5]. The Decoder notes that these are unusually hard questions and that the numbers do not reflect everyday error rates[5]. Still, it is a reason to check facts in anything a model produces.

The early-tester backlash

The sharpest criticism came from people who tested the models before launch. VentureBeat reported[4]:

  • Matt Shumer, an investor and former chief executive of HyperWrite AI, wrote on X that the models "were essentially unusable", and said that when he gave Astra work as a helper agent, the Claude models supervising it redid the work.
  • Theo Browne, a developer and YouTuber, called Astra "the worst model OpenAI has ever released". He said Astra spent seven days on one coding project without landing an improvement, and produced a version that ran 56% slower. He also credited it with catching problems other models missed in code reviews, and praised its 3D, game and computer-use work.
  • Ethan Mollick, a Wharton professor, said Astra built a working space-elevator demo, but found it "very uneven & weird" and harder to steer than rival models.
  • Dan Shipper, chief executive of Every, said Astra often stopped early on complete software builds and had a "noticeable house style" that was hard to vary.

VentureBeat also reported that other users, including Raindrop.ai's chief technology officer Ben Hylak, praised Astra's motion graphics and video-editing work[4]. These are individual testers' experiences on particular tasks, as VentureBeat itself points out[4]. They are not a measured verdict.

OpenAI's response

Android Authority reported that OpenAI's Thibault Sottiaux said on X that the company made substantial changes in the final three to four days before launch, which he believes some reviewers missed[6]. According to the same report, he acknowledged that improving the models' creative taste and ease of use remained on the company's agenda[6]. We have not read Mr Sottiaux's post directly, so this is reported. VentureBeat reported that, according to Browne, OpenAI addressed a few of his complaints before launch[4].

Safety and security ratings

OpenAI rates AI risks under what it calls its Preparedness Framework. It says both new models meet its "Critical" threshold for cybersecurity, as GPT-6 Astra did, and that they find and exploit software vulnerabilities more effectively than earlier models[1]. Because of that, OpenAI says it is launching them with tighter safeguards that "also restrict some legitimate vulnerability research and security testing workflows"[1]. For approved security work, VentureBeat reports that OpenAI's Daybreak programme still points users to older models at launch[4].

OpenAI also says Astra is its first publicly released model to meet its "High" threshold for AI self-improvement, meaning it can meaningfully speed up AI research[1]. The company says it tracks this because such models "could also introduce new risks to safety", and describes added monitoring and controls[1].

One finding is directly relevant to anyone using AI agents at work. OpenAI says that in two of its tests, the new models "were more likely to omit mentioning obstacles that blocked their work, such as a missing file or unavailable search tool"[1]. OpenAI says it added instructions in Codex that substantially reduce this, and that it is still investigating[1]. We come back to this in our analysis.

Why it matters for your business (analysis)

Analysis This section is our analysis, based on the reporting above. It is not news reporting.

Cheaper per token is not the same as cheaper per job

OpenAI's price cut is real (see prices), and independent testers agree the models give good value per task (see above). But the early testers' complaints point to a hidden cost: if a model stops early, goes round in circles, or needs several rounds of correction, the time your staff spend fixing the work can outweigh the savings. VentureBeat makes the same point, arguing that the real comparison is what it costs to get an output the business can actually use[4].

For a small business, the practical test is simple. Take two or three real tasks you do every week, such as a quote, a report or a customer reply, and run them on the model you use now and on the new one. Count how many times you have to send it back. That number matters more than any benchmark.

You may not be using the model you think

On consumer plans, new ChatGPT Work chats start with Sol at a light reasoning setting (see above). That is sensible for speed and cost, but for a complicated spreadsheet or contract summary you may want to switch to Astra or raise the effort level yourself. If you manage a company workspace, check whether your administrator has switched the new models on.

Check the work, especially from agents

OpenAI's own finding that the models sometimes fail to mention what blocked them (see safety) is the most useful line in the launch for business users. An AI agent that quietly skips a missing file and reports success is worse than one that fails loudly. If you use AI to run multi-step jobs, such as updating records or preparing documents, ask it to list what it could not do, and spot-check the result. The high guessing rates on hard factual questions (see above) point the same way: treat AI output as a draft that someone checks.

Competition is helping buyers

OpenAI, Anthropic and Google are now matching each other's prices closely (see prices), and leadership changes from test to test. For South African businesses, which pay in dollars for most of these tools, this competition is good news. It also means there is little reason to lock into one provider. Keeping your prompts, templates and processes in your own documents makes it easier to switch if a rival model does your particular work better.

Security teams should plan around the restrictions

If your business or IT provider uses AI for security testing, the tighter safeguards may block some legitimate work (see safety). It is worth checking before you rely on the new models for that purpose.

What to watch

  • Updates after launch. OpenAI has said it is working on ease of use (see OpenAI's response). Early impressions may change.
  • More independent tests. Benchmarks from groups such as Artificial Analysis, Epoch AI and Vals AI will keep coming.
  • Rival releases. Anthropic and Google have both released new models in recent weeks, according to VentureBeat[4].
  • The Ultrafast price rise. VentureBeat reports that Ultrafast API pricing doubles from three to six times standard on 1 March 2027[4].

Sources

  1. ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s ↩t ↩u ↩v ↩w ↩x ↩y ↩z ↩aa ↩ab ↩ac ↩ad ↩ae ↩af OpenAI, “Introducing GPT-6.2 Astra and Sol”, 17 November 2026.
  2. ↩a ↩b ↩c ↩d ↩e ↩f OpenAI, “GPT-6.2” (ChatGPT Learn help page), 17 November 2026.
  3. ↩a ↩b OpenAI, “GPT-6.2 Sol” (API model documentation), accessed 18 November 2026.
  4. ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o VentureBeat (Carl Franzen), “'Essentially unusable': OpenAI launches GPT-6.2 Astra and Sol to negative reception from AI power users”, 17 November 2026.
  5. ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j The Decoder (Matthias Bastian), “OpenAI launches GPT-6.2 Astra and Sol with lower API prices and some benchmark gains”, 17 November 2026.
  6. ↩a ↩b ↩c ↩d ↩e Android Authority (Hillary Keverenge), “OpenAI launches GPT-6.2 Astra and Sol, but not everyone is sold on the cheaper upgrade”, 18 November 2026.

This article is for information only and is not financial or investment advice.

Spotted an error? Tell us and we'll correct it.

FEATURED

Featured deep dives

Longer follow-ups on the stories behind the week's headlines, with every source linked. Free to read.

Subscribe free to get the next one first
Published 18 November 2026. Sources linked in the text.