Models · Hardware · Providers
Ask anything about running AI —
then see the data behind the answer.
Every answer is grounded in our live database of 487 models, 71 devices from data-centre GPUs to phones, and 1,989 provider price listings. We re-read each listing on its own schedule — the newest 6 hours ago, the oldest 3 months ago.
no sign-in · answers grounded in live catalogue data
01487 entries
Models
Every open model, fully specced.
VRAM per quant, estimated tokens/sec, benchmarks, licenses — synced from primary sources daily.
0271 entries
Hardware
GPUs, Macs and phones, ranked for AI.
What fits, how fast it runs, and what to buy next — from an RTX 5090 down to the phone in your pocket.
0390 providers
Providers
Hosted inference, priced daily.
1989 live offers with per-token pricing, served quants, context limits and privacy terms.
Latest releases & changes
All news →Host Mancer 2 raised gpt-oss-120b input pricing by 10%6 hours agoKimi K3 repriced across 2 hosts: Wafer (fast tier) input and cache read down 5%, Relace input up 1%6 hours agoHost Inceptron raised Kimi K2.6 cache-read pricing by 34%6 hours agoHost Baidu repriced DeepSeek V4 Pro: output down 15%, input up 28%6 hours agoGLM 5.3 Flash repriced across 4 hosts: OpenInference output down 31%, Sail Research output up 300%12 hours agoGLM 5.3 repriced across 2 InferenceNet listings: InferenceNet (US region) output down 23%, input and cache read up 30%12 hours ago
