Larry Dignan

Writer, which specializes in marketing and revenue AI workflows, released its Palmyra X6 flagship model and upgrades to its Writer Agent harness. A few points stuck out in the announcement.

• Writer is targeting token spending. The company said Writer Agent operates at a 52% lower cost with 48% improvement in speed when using Palmyra X6.
• Palmyra X6 was trained on top of GLM-5.2.
• Writer benchmarked its latest model and harness compared to frontier models and outperformed.
• Palmyra X6 completes tasks in 26 seconds on average, generates 82 tokens per second and can work unattended for up to 8 hours.

MongoDB launched a set of tools that bring automated embeddings to its Atlas platform via Voyage AI. The company said the Atlas Embedding and Reranking API, voyage-code-4, and vector search in Atlas Stream Processing all are designed to improve retrieval accuracy. MongoDB also announced Atlas Managed MCP Server, a fully hosted service that connects Claude Code, Codex, Grok Build, and Devin to MongoDB Atlas.

Oracle said Oracle Exadata Database Service on Exascale Infrastructure on Oracle AI Database@AWS is now generally available. Oracle and AWS said they have also committed to a new long-term collaboration pact to continue investing in Oracle AI Database@AWS.

Mistral said it will expand its platform to open models beyond its own. Mistral's platform will support third-pary open models starting with Z.ai's GLM-5.2. The third-party open models will run on Mistral's infrastructure, regional controls and service commitment used by its own models.

The effort was outlined in a post detailing Mistral's sovereign AI plans. Notably, Mistral said its Regional Endpoints are generally available. The service lets customers choose whether their inference runs in the US or Europe. The company also said its Mistral Priority Tier is now in public preview.

Cerebras CEO Andrew Feldman had interesting quotes on the company's second quarter earnings call. It's the AI equivalent of the "speed kills" phrase in sports. He said:

• "The market is realizing that speed is not a benchmark item. Speed changes user engagement, it changes agentic performance, and it changes AI productivity. Fast inference unlocks new applications and new markets."
• "On the capabilities front, in the second quarter, we delivered support for OpenAI's GPT-5.6 Sol, the largest and most capable of the frontier models. In fact, Cerebras serves 5.6 Sol at a speed that is 10x faster. With GPT-5.6 Sol, this lays to rest any of the remaining concerns regarding our ability to support large frontier models."
• "Speed is critical for user experience. Throughput is critical for inference economics."
• "Cerebras wants more throughput without giving up speed. Herein is the strength of our disaggregated solution. It delivers Cerebras speed with 5x higher throughput. Increasing throughput by 5x while keeping our industry-leading speed has a profound impact on the economics of token generation. It means up to 5x as many high-speed, high-value tokens are made by each Cerebras system. More tokens per system at lower cost means more revenue and more gross margin."

In the big picture, Cerebras' quarter was a solid building block. Wall Street wanted more an shares were down 16% premarket.

Infleqtion reported second quarter revenue of $12.6 million, up 116% from a year ago. The quantum computing company projected 2026 revenue of about $43 million. CEO Matt Kinsella said "the pace of quantum commercialization is accelerating" and "governments are putting dates and dollars behind quantum."

The company said its second quarter net loss was $25.47 million.

SpaceXAI launched Grok 4.6, a fast follow to Grok 4.5. Grok 4.6 is built for long-running tasks and more complicated tasks. The model appears to be on par with OpenAI's GPT-5.6 Sol. Grok 4.6 is available in Cursor and Grok build.

You probably won't see this in the Pixel 11 ads from Google, but the company raised prices of its devices, downsized the memory and limited the Google One freebies.

The Pixel 11 launch featured a bit of inflation as well as shrinkflation. Google's Pixel launch's included starting prices for the Pixel 11 all $100 more than the equivalent Pixel 10 devices. The Pixel Watch 5 starts at $50 more than the previous model.

Google's device prices run from $399 (Pixel Watch 5) to Pixel 11 at $899 to Pixel 11 Pro at $1,099 to the Pixel 11 Pro Fold at $1,899. The base versions of the Pixel 11 Pro and Pro XL have 256GB starting storage, double the Pixel 10 versions. But RAM in those base devices are now 12GB compared to 16GB in the Pixel 10.

Meanwhile, Google cut the Google One AI Pro trial bundle with the Pixel 11 Pro and other devices from 12 months to six months. If you buy the Pixel 11 there is no free trial at all.

There were a few upgrades with Google Tensor G6 processor, slight camera upgrades and more durability.

In the end, Google just confirmed what we all know: 2026 is not the year to upgrade your personal devices due to soaring component costs. Thanks AI.

Super Micro reported fourth quarter net income of $1.18 billion, or $1.62 a share, on revenue of $11.1 billion, up from $5.8 billion a year ago. Super Micro preannounced the quarter.

For fiscal 2026, Super Micro reported net income of $2.2 billion on revenue of $39.1 billion. As for the outlook, Super Micro projected first quarter sales of $14.5 billion and $15.5 billion with non-GAAP earnings of $1.01 a share to $1.10 a share. Fiscal 2027 revenue will be between $65 billion to $72 billion.

Brad Lightcap, OpenAI's former chief operating officer, said he will be leaving the company to "start something new." Lightcap joined OpenAI in 2018. His most recent role was chief operating officer. That gig ended in April and he was given a role focused on special projects. OpenAI has been reshuffling its executives.

Epicor said Vaibhav Vohra will become CEO Oct. 1. Current CEO Steve Murphy will step down at the end of Epicor's fiscal year and remain on the Board of Directors.

Vohra has been Epicor's President since October 2024. He joined Epicor in 2021 as Chief Product Officer.

The Wall Street Journal has a story detailing the tough investor questions Anthropic is seeing ahead of its IPO, which should be this Fall. According to The Journal:

"Company executives have played down the impact of Chinese competition in meetings, telling investors that Anthropic is hyperfocused on offering cutting-edge AI models, the people said. Chief Executive Dario Amodei and other top U.S. AI leaders have publicly said that most users want the most intelligent AI systems available at any given time, suggesting that Chinese systems are less of a threat since their capabilities generally trail those of top AI models by at least a few months."

A business that depends on being a few months (or weeks) ahead of open models is one of two companies driving the AI economy. There are no details on pricing of shares or the financials yet. I can't wait to see the numbers.

• Anthropic: Opposed to open model ban, wants safety testing for all models
• My internal monologue reading Anthropic's big idea posts