SubQ Snap gives the first 10,000 signups 100M free input tokens, worth $5 in API credit
Subquadratic just released its SubQ Snap decision model in public preview, and the first 10,000 people to sign up get 100 million free input tokens, worth $5 in API credit. Snap only bills input at $0.05 per million tokens, output is free, and a single request can take up to 384K tokens of text and images.
The briefing
Subquadratic is an AI research company focused on making model algorithms scale more efficiently. On October 12 it released SubQ Snap, a decision model now open in public preview, and the first 10,000 signups get free credit to try it.
What you get
The first 10,000 users who sign up for the SubQ Console get 100 million free input tokens. Snap costs $0.05 per million input tokens, so the grant shows up as $5 of balance in the console. It's first come, first served, and once capacity is reached new users may be put on a waitlist.
What you can use
Instead of writing open-ended text, Snap returns a probability for each answer you define, which suits the routine calls inside an agent loop: routing, picking a tool, checking whether a test passed, or deciding which step needs a frontier model. Only input is billed, and output is free.
Each request can hold up to 384K tokens of mixed text and images, and a 100K-token request comes back in about a second.
Specs and public benchmarks
- On the Decision Index 0.2.1 public suite, Snap scores 61.7 versus 57.9 for Jev.
- Across 16 enterprise workflow benchmarks, Snap averages 67.5 versus 61.6 for Jev and leads on 13 of them.
- On Decision LCR multi-document reasoning, Snap scores 74.0 reading everything in one shot, while a chunked Jev scores 48.0.
How to claim
- Sign up at the SubQ Console. The free credit lands right after you register.
- Create an API key in the console and start calling SubQ Snap.
Details: SubQ Snap announcement and pricing page.