LIVE
ANTHROPICOpus 4.7 benchmarks published2m ago
CLAUDEOK142ms
OPUS 4.7$15 / $75per Mtok
CHATGPTOK89ms
HACKERNEWSWhy has not AI improved design quality the way it improved dev speed?14m ago
MMLU-PROleader Opus 4.788.4
GEMINIDEGRADED312ms
MISTRALMistral Medium 3 released6m ago
GPT-4o$5 / $15per Mtok
ARXIVCompositional reasoning in LRMs22m ago
BEDROCKOK178ms
GEMINI 2.5$3.50 / $10.50per Mtok
THE VERGEFrontier Model Forum expansion announced38m ago
SWE-BENCHleader Claude Opus 4.772.1%
MISTRALOK104ms
ANTHROPICOpus 4.7 benchmarks published2m ago
CLAUDEOK142ms
OPUS 4.7$15 / $75per Mtok
CHATGPTOK89ms
HACKERNEWSWhy has not AI improved design quality the way it improved dev speed?14m ago
MMLU-PROleader Opus 4.788.4
GEMINIDEGRADED312ms
MISTRALMistral Medium 3 released6m ago
GPT-4o$5 / $15per Mtok
ARXIVCompositional reasoning in LRMs22m ago
BEDROCKOK178ms
GEMINI 2.5$3.50 / $10.50per Mtok
THE VERGEFrontier Model Forum expansion announced38m ago
SWE-BENCHleader Claude Opus 4.772.1%
MISTRALOK104ms

All Catalogs

Every TensorFeed catalog and signal surface in one place. 37 free, machine-readable agent-shaped data sources covering compute, models, training, agent infrastructure, evaluation, live signals, discovery, and industry context.

For programmatic discovery, agents should use /api/meta, /llms.txt, or /openapi.json.

Compute

The chips, the rentals, what training costs, where to run it.

Models

Every model worth picking from, by tier.

Training & Customization

Datasets, fine-tuning paths, what training actually costs.

Agent Infrastructure

Frameworks, harnesses, MCP, the APIs agents wire in.

Evaluation & Safety

Benchmarks, registries, what the labs say about their models, real production usage.

Live Signals

Real-time operational data we measure ourselves.

Discovery & Ecosystem

Where to play, where to find more, where the conversation lives.

Industry Context

Money flowing in. Rules being written.

For Developers

Where to start if you are wiring TensorFeed into something.

Missing something? The catalog grows weekly. Editorial cadence is roughly weekly for fast-changing surfaces (funding, leaderboards, attention) and on redeploy for slower ones (frameworks, hardware specs). Send feedback to [email protected].