Launches on 2026-09-17

A blended model that runs several models on every request and keeps the best answer — 262K context, $2.50/$7.50 per million tokens.

Merchant-of-record payments for software and digital products: one rate of 3.9% plus 40 cents, with tax, payouts and affiliates included.

Cloud Macs and Linux machines that match your CI exactly, for developers and for coding agents — start in seconds, archive over the weekend.

Sends synthetic callers at a voice agent over real phone calls — accents, interruptions, background noise and prompt-injection attempts — and scores what breaks.

Testing and evaluation for MCP servers: see how a server behaves inside ChatGPT, Claude and Copilot, and gate every pull request on it.

Watches competitors across search, ads, social and the AI assistants, and sends one Monday briefing on what changed and what to do about it.

Upload a pitch deck, get matched to venture firms whose thesis fits, and let Pitchfire send the introduction — free.

Brief an AI by phone, text or email and get back an editable PowerPoint and a narrated video — $29 a deck, no subscription.

One asynchronous API for more than 50 image and video generation models, billed per second of output.

A 27-billion-parameter model compressed to 5.9 GB of ternary weights, released under Apache 2.0 — small enough to run on a laptop.