Meta launched Muse Spark 1.1 with aggressive developer pricing, OpenAI broadly released its GPT-5.6 series with a 54% efficiency claim, and Anthropic introduced a usage-analytics feature for Claude — all on the same day.
Three of the largest players in artificial intelligence rolled out new products on Thursday, each targeting a market where the competitive stakes are rising rapidly: agentic coding and developer adoption.
Meta introduced Muse Spark 1.1, the first significant update to the model it launched in April under AI chief Alexandr Wang. OpenAI made its GPT-5.6 Sol, Terra, and Luna models broadly available after an initial limited release. Anthropic, meanwhile, unveiled a user analytics feature called Reflect for its Claude chatbot.
Wang, who leads Meta's Meta Superintelligence Labs unit, described Muse Spark 1.1 as the company's "strongest model for agentic and coding work yet" in an interview with CNBC. The original Muse Spark, released in April, was restricted to select partners via a private API preview. The updated version opens access through a developer portal as part of a public preview, with a waitlist for new users.
Meta said it is currently limiting API access to its own properties and will not make it available on third-party platforms such as OpenRouter.
Wang characterized the pricing as "very aggressive and attractive" relative to comparable offerings from Anthropic and OpenAI. New API accounts will receive $20 in free credits. After that, the company will charge $1.25 per million input tokens and $4.25 per million output tokens. "The goal is to really have attractive pricing that scales with immense consumption usage," Wang said.
Wang added that Meta is training a more powerful model, internally code-named Watermelon, though he declined to give a release timeline. He also said a variant of Muse Spark intended for the open-source community is in development, without specifying when it would be released.
The Muse Spark launch is part of a broader effort by Meta to diversify its revenue beyond advertising. CEO Mark Zuckerberg has faced pressure from investors to demonstrate returns on the company's growing infrastructure spending — outlays that parallel those of cloud competitors, though Meta does not yet operate a commercial cloud business.
On the OpenAI side, CEO Sam Altman told CNBC that GPT-5.6 Sol is 54% more token efficient on agentic coding tasks and is "as good or better" than competing models. "Every enterprise now is thinking about spend and the value they're getting in exchange for AI, and this is what we really want to do," Altman said.
OpenAI had initially limited the GPT-5.6 series to a small group of trusted partners at the U.S. government's request. Altman said the company worked with Commerce Secretary Howard Lutnick, Treasury Secretary Scott Bessent, and U.S. National Cyber Director Sean Cairncross during the approval process, describing the relationship as a "collaborative back and forth" in which government officials ran tests and flagged concerns.
"If you want broad access, which we do, and you have powerful models, you really want to be able to be confident in your safety claims," Altman said.
OpenAI is currently valued at $852 billion by private investors. Altman declined to say whether the company would pursue an IPO this year, saying only, "I don't know."
Anthropic's contribution to the day's announcements was Reflect, a built-in dashboard for Claude that lets users track and visualize their AI usage patterns — including topics discussed, overall activity, and the types of tasks they rely on Claude to handle. The feature also allows users to set quiet hours and schedule reminders to take breaks. Anthropic said data shown in Reflect is not used for other purposes, and that conversations connected to health integration tools are excluded from insights entirely. Claude Reflect is available in beta for Free, Pro, and Max users who have memory enabled.
The simultaneous releases on Thursday underscore how quickly the competition for developer loyalty has tightened in 2026, with pricing, efficiency benchmarks, and workflow integration all emerging as primary battlegrounds alongside raw model capability.
TechEchelon Staff bylines are produced collectively by the newsroom for short, breaking, and wire-style coverage. Longer-form reporting is published under the responsible reporter's name.
More from the Staff →Meta Platforms announced plans to build a 1 gigawatt data center in Alberta, Canada — its first major facility in the country — at an estimated cost of $9 billion, as the company races to expand AI infrastructure capacity.
The European Commission adopted a nine-measure cybersecurity and AI action plan on July 7, aiming to reduce the bloc's dependence on foreign frontier AI systems after U.S. export restrictions temporarily cut off EU customers from models including OpenAI's GPT-5.6.
OpenAI will publicly release its GPT-5.6 Sol, Terra, and Luna models on Thursday after the U.S. Department of Commerce cleared the company to proceed, ending a roughly two-week restricted rollout requested by the government.
Written for readers who already know the basics — markets, AI, and the policy decisions that shape both.