New playbook: when your own AI model pays. Read
Grab (superapp, southeast asia, ~11,000 employees) reported this on 24 Jul 2026 in company self-report (source). It belongs to the models and routing layer.
The GrabGPT Gateway, an OpenAI-compatible endpoint that fronts every model call in the company (five providers); the LLM-Kit framework resolves the endpoint per environment and data tier so platform teams can change providers, fallbacks, budgets and cost attribution without app changes.
500+ services and 50+ MCP servers on the framework; billions of tokens per month through one gateway (no cost figure published).
The number covers scale and coverage metrics only.
“If moving from OpenAI to Anthropic... takes more than a few config changes, technical debt is already slowing you down,” Grab said.
One gateway picks the AI model for each task and records what it costs.
Up to 90%
lower AI cost, AT&T says
An internal AI Gateway that, turn by turn, weighs speed, cost and expected quality and routes each prompt to the most cost-effective model, switching models mid-session, plus a library of post-trained open telco models (OTel 2.0) that run on-prem or in cloud (source).
<8%
of closed-model cost per transaction
A blended model portfolio (in-house compact models, open-source models post-trained on Pinterest data in its own secure cloud, closed models only in 'very limited cases') behind a centralized routing layer that sends routine work to lighter models; being extended to internal engineering and sales tools in H2 2026 (source).
52%
lower cost per agent session since the June peak
A cost equation and weekly metric stack for its agentic 'software factory' (70%+ of PRs agent-attributed), with benchmark-driven Pareto model selection (incl. open-weight models), cheaper default subagent models, Medium reasoning effort default, 400K-token compaction cap and prompt-cache strategy (source).