L40S LLM benchmark
No signed L40S submissions yet. Check the configuration guidance below, or contribute a reproducible measurement from your own machine.
Specs vendor specs, not measured
- VRAM
- 48 GB
Plan memory before choosing a model
The catalog lists 48 GB of VRAM for this hardware entry. Check the capacity of your actual card. Model weights, context cache and runtime overhead must share the available memory. Capacity alone does not establish which model will fit or how fast it will run. Use the memory-fit calculator for your model and context as an estimate, then verify with the intended runtime. No signed L40S measurement is available yet.
Contribute a signed L40S benchmark
Run on YOUR hardware to populate this page: pipx install llm-speed && llm-speed bench
$ pipx install llm-speed && llm-speed bench
Community folklore on L40S
2 unverified claims extracted from Reddit/HN comments. Lower trust than signed runs above; every row links to the source.
- communityconfidence 60%
15.00tok/s — gpt-oss 120B on L40S via ollama
our signed data: L40S · gpt-oss 120B
“.1 t/s and 1173 t/s eval and prompt eval respectively on this system. Even on an L40 which can load it totally into VRAM, R1-70B only hits 15t/s eval. (gpt-oss 120B doesn't run reliably on my single L40 and gets much slower when it does manage to run partially in VRAM on that s…”
- communityconfidence 60%
4.10tok/s — deepseek-R1 on L40S via ollama
our signed data: L40S · deepseek-R1
“ompt eval using GPT-OSS 120B in ollama, with no special hackery. For comparison, the much smaller deepseek-R1 70B takes the same prompt at 4.1 t/s and 1173 t/s eval and prompt eval respectively on this system. Even on an L40 which can load it totally into VRAM, R1-70B only hits…”
Request a run on L40S
No measured run yet for these pairs. Vote for the one you want; the most-requested runs go first. See all requests.
We never sell or share your email. Privacy