Independent Pricing & Quality Intelligence for AI Infrastructure
Vector databases, LLM APIs, inference platforms, GPU clouds, agent tooling. Consumption-level prices extracted weekly from public pages, logged with a SHA-256 hash, chained per vendor, and archived — a financial time-series that cannot be retroactively faked or fast-forwarded.
The instruments
The Tipping Point
Per-token API vs renting raw GPUs — the exact volume where the lines cross, on your throughput numbers.
Public ledgerAI Pricing Pulse
A terminal feed of every observed change — rate limits halved, token prices moved — each entry carrying its verification hash.
TrustHow We Score & What We Share
The exact 12-signal formula, the hash-chain spec, and our advertising disclosure. Vendors cannot pay to change a number.
$9 / $29 moLedger Pro — full history
The public Pulse shows a rolling window. Pro unlocks 5-year CSV exports and API access to every verified record.
Browse by category
LLM APIs
Input vs output token rates, cache-read shadow pricing, context windows.
10 vendorsVector Databases
$/GB storage, per-million-dimension pricing, read/write QPS.
10 vendorsInference Platforms
Serverless token rates and dedicated-endpoint economics.
6 vendorsGPU Clouds
H100/A100 hourly rates — the denominator of every self-hosting decision.
6 vendorsAgent Tooling
Orchestration platforms and the token overhead agents really add.
6 vendorsLLM Observability
Tracing and eval platforms — the meter on the meter.