Trendyol LLM Asure 12B on NVIDIA GeForce RTX 5080
Measured 3 mo ago.
Measurement
- tok/s
- 61.5
- TTFT
- 323 ms
- VRAM used
- —
- RAM used
- —
- Power
- —
- Quant
- Q4_K_M
- Context
- 8K
- Run date
- 2026-05-27
- Source
- owner
- Type
- measured
V36.52 rigor detail
Protocol →- Cold-start decode
- 61.58 tok/sTTFT 6510 ms
- Steady-state median
- 61.52 tok/sP5 61.5 · P95 61.6
- Runs captured
- 3
- Scenario
- Single-stream
3-run measured Ollama API median; variance 0.1%; raw logs in public Gist
Evidence
What this row provides for independent review. Missing fields lower confidence; they are shown explicitly instead of hidden.
- Source link
- Open source
- Evidence manifest
- Open manifest
- Command
- Available
- Raw logs
- Not provided
- Raw results
- 1 files
- Log hash
- 73c7fe1e1310...
- Operator
- fred-oline
- Runtime
- ollama version is 0.24.0
- Driver
- NVIDIA 595.97
- CUDA / ROCm / Metal
- Not provided
- OS
- Windows
- Run count
- 3
- Raw stats
- Available
- Environment notes
- Not provided
Why this confidence tier?
Confidence is rule-based. Every factor below contributed to the tier. We never expose a single numeric score; the tier label is auditable through this explanation alone.
- +Measured by RunLocalAI editorial
- Reproduce this benchmark →An independent reproduction with matching numbers lifts the tier and reduces single-source risk.
- Read the confidence methodology →Full editorial standards for tiering.
- Why we don't use percentages →Tier labels — auditable, no opaque score.
Cohort intelligence
How this measurement compares to the rest of the corpus. Only comparable rows (same model + hardware first, with relaxations labelled) are used. We never average across runtimes or quant formats unless explicitly told to.
Same model + hardware, different runtime
2 matching rowsVariance here is pure runtime / version drift. Wide spread suggests a runtime regression candidate worth investigating.
- 79.1 tok/srtx-5080ollama-local-apiunknownEditorial
- 82.0 tok/srtx-5080ollama version is 0.24.0Q4_K_MEditorial
Same model + hardware, different quant
1 matching rowHow much speed you trade for memory. Useful when the operator is deciding whether to drop to a smaller quant.
- 79.1 tok/srtx-5080ollama-local-apiunknownEditorial
Same model, different hardware
1 matching rowWhat this model looks like on adjacent hardware. Drives the 'should I upgrade?' question.
- 43.4 tok/srtx-3080-16gb-mobileollama version is 0.24.0Q4_K_MEditorial
Same hardware, different model
8 matching rowsWhat else this rig can run at the same quant bucket.
- 79.0 tok/srtx-5080ollama version is 0.24.0Q4_K_MEditorial
- 101.1 tok/srtx-5080ollama-local-apiQ4_K_MEditorial
- 133.4 tok/srtx-5080ollama-local-apiQ4_K_MEditorial
- 133.6 tok/srtx-5080ollama-local-apiQ4_K_MEditorial
- 135.6 tok/srtx-5080ollama version is 0.24.0Q4_K_MEditorial
- +3 more
Reproduce this benchmark
Got the same model + hardware combo? Run the same measurement and submit your numbers. We'll pre-fill model, hardware, quant, and context — you just add your tok/s, VRAM, runtime version. If your numbers match within ±15%, this benchmark gets a confidence lift and a reproduction badge.
Related
Drill into the entity pages for this measurement.
Cite or export
Reference this benchmark in your work. Multiple formats; CC-BY attribution required.
Cite this benchmark or paste it into a README. Copy-to-clipboard; license is CC-BY-4.0 (attribution to RunLocalAI required).
<a href="https://www.runlocalai.co/benchmarks/345" rel="noopener">RunLocalAI: Trendyol LLM Asure 12B on NVIDIA GeForce RTX 5080 — 61.5 tok/s</a>
Next recommended step
Got the same model + hardware? Run it and submit your numbers — successful reproductions lift this benchmark's confidence tier.