UncoverAIFollow

The Enterprise AI Briefing: August 2026 Edition

(Synthesizing the Market Signals of July 2026)

01

Model

GPT-5.6 Goes Generally Available in Three Tiers

The Signal

On July 9, OpenAI made the GPT-5.6 family generally available after a limited preview: Sol, the flagship; Terra, a balanced model for everyday work; and Luna, the most cost-efficient of the three. At general availability OpenAI priced Sol at $5 per million input tokens and $30 per million output tokens, Terra at $2.50 and $15, and Luna at $1 and $6.

The same release introduced ultra, which coordinates four agents in parallel by default, and max, which gives the model more time to reason than the xhigh setting. On July 30, OpenAI reduced Luna's price by 80 percent and Terra's by 20 percent.

Reuters reported that GPT-5.6 also debuted that Thursday as the model behind ChatGPT Work.

The Noise

Leaderboard gaps between GPT-5.6 Sol and Claude Fable 5.

The Enterprise Impact

A single generation now ships as three priced tiers, and ultra spends more tokens by running agents in parallel. Procurement has to price Sol, Terra, and Luna separately, and treat the July 30 cuts as a change to Terra and Luna only, not to Sol.

02

Model

Claude Opus 5 Arrives at the Prior Opus Price

The Signal

On July 24, Anthropic made Claude Opus 5 available, describing it as close to the frontier intelligence of Claude Fable 5 at half the price. It is the new default on Claude Max and the strongest model on Claude Pro.

Developers reach it as claude-opus-5 at $5 per million input tokens and $25 per million output tokens, the same rates as Opus 4.8. Fast mode runs at about 2.5 times the default speed and costs twice the base price.

Anthropic said the model remains behind Mythos 5 on cybersecurity and does not advance the frontier in biology or offensive cyber.

The Noise

Early-access anecdotes about weekend coding sessions.

The Enterprise Impact

Buyers get an Opus-class default at the previous Opus price, with a fast mode that doubles the bill. Security teams still cannot treat Opus 5 as a stand-in for Mythos 5 on exploit work, and the Cyber Verification Program remains the path to the less restricted variant.

03

Acquisition

SAP Closes Prior Labs and Commits More Than €1 Billion

The Signal

On July 17, SAP completed its acquisition of Prior Labs, which it calls the pioneer of tabular foundation models. Prior Labs continues as an independent entity.

SAP committed to investing more than €1 billion over the next four years to scale that lab for the structured data underneath business systems. Prior Labs said regulatory approvals were secured, that its models stay open, and that it is backed by more than €1 billion.

The signing was announced in May. July is the close, and neither company stated a purchase price.

The Noise

Reading the four-year investment commitment as the price SAP paid.

The Enterprise Impact

Tabular models for ERP-style data now sit inside SAP as a named lab, not as a side experiment. Customers should expect those models to show up in the suite, and they should treat the open-model commitment as a term to recheck as the lab scales.

04

Funding

Databricks Signs a Term Sheet at a $188 Billion Valuation

The Signal

On July 16, Databricks said it had signed a term sheet for a strategic round at a $188 billion valuation, led by existing investor Coatue and open to new and existing investors, with a close expected later that summer. The company said the capital would go to Unity AI Gateway, Genie, and Lakebase.

Databricks did not state the size of the round. The Wall Street Journal, citing people familiar with the matter, reported that Coatue was leading a $3 billion investment at that valuation, and Reuters attributed the $3 billion figure to the Journal.

The Noise

Treating the Journal's $3 billion, or a signed term sheet, as closed capital Databricks has confirmed.

The Enterprise Impact

The round, if it closes, funds the layer enterprises use to choose models and govern their cost. Architecture plans that assume the money is already in the bank are ahead of the company's own timetable.

05

Partnership

AMD and Anthropic Commit Up to 2 Gigawatts and $5 Billion

The Signal

On July 22, AMD and Anthropic announced that Anthropic will deploy up to 2 gigawatts of AMD Instinct MI450 Series GPUs in AMD Helios racks, with the first gigawatt beginning in the first half of 2027. Those racks pair MI455X GPUs with EPYC Venice CPUs, Pensando networking, and ROCm software.

The companies will use Claude to optimize workloads for Instinct GPUs and to accelerate ROCm, and AMD will broadly adopt Claude across its engineering and product teams. AMD committed to a strategic equity investment of up to $5 billion in Anthropic.

The Noise

Reading "up to" as gigawatts and equity that have already been delivered.

The Enterprise Impact

Claude's capacity plan now includes a second silicon path, and the first gigawatt is not scheduled until the first half of 2027. Enterprises standardizing on Claude should not assume AMD capacity is online in 2026, and the equity commitment is a future investment, not a closed round.

  • AMD22 July 2026