Fit receipt — a fixed address for one verdict

gpt-oss-120b on RTX 4090

WON'T FIT ✗

Conditions: Q4_K_M · 8K context · KV F16 · Windows + display reserve

weights	KV cache	overhead	reserve	total	short by
66.6 GB	0.3 GB	8.3 GB	2.0 GB	77.1 / 24 GB	53.1 GB

→ A single A100 80GB would fit it.

Replay: npx fitllm "gpt-oss-120b" --gpu "RTX 4090" --quant Q4_K_M --ctx 8192 --kv 16 · JSON · interactive

live fit badge — embed: ![fits](https://img.shields.io/endpoint?url=https%3A%2F%2Ffitllm.run%2Fapi%2Fbadge%3Fmodel%3Dgpt-oss-120b%26gpu%3DRTX%2B4090%26quant%3DQ4_K_M%26env%3Dwindows-display%26ctx%3D8192%26kv%3D16)

Computed by the open fitllm-engine (MIT) from official config.json values — estimates; runtime varies. Ran it for real? Challenge this prediction — measured reports calibrate the engine for everyone.

engine 2.21.4 · revision 2dca33db8c822bb2348949b5c54797d86a869202 · release manifest