Your GPU. Exact local AI configurations. Real execution evidence.
Start from the GPU you already own: see what RigFact actually tested, what happened, and what is still unknown.
Models relevant to local AI in 2026
Current means relevant and available as of 2026 — not necessarily released in 2026. Older releases stay listed while they remain current choices, and a recent release date alone proves nothing. A release date is provenance, not a currency verdict.
Catalog presence means the model exists and is discoverable — it does NOT mean RigFact tested it, that it fits your GPU, or that we recommend it.
Server-class models are discoverable but are not a single consumer-card target: GLM-5.3 · Kimi K3 · DeepSeek V4-Pro · Qwen3.5 397B · Llama 4 Scout · gpt-oss 120B · Mistral Medium 3.5.
18 pre-2026 entries are preserved as LEGACY, plus 1 publisher-deprecated model kept discoverable with the publisher's notice intact — never erased, never silently promoted to current. Browse the model catalog →
Advanced tools
For deeper technical work, once you know what you are looking for.
- VRAM Calculator — will this model fit in my GPU memory?
- Recipe Planner — plan an exact model and configuration you already have in mind.
- Evidence Compare — compare captured runs side by side.
- Model Catalog — the model reference inventory and its provenance.