Speed sweep
qwen3-6-35b-a3b-ud-iq4-nl-rtx-3090-24gb-llama-cpp-tp1-sweep
qwen3-6-35b-a3b-ud-iq4-nl-rtx-3090-24gb-llama-cpp-tp1-sweepRecord
- Measured
- 2026-05-13T18:15:37.846Z
- Points
- 1
Measured speed
| Concurrency | Context | Prefill | Decode | TTFT ms | Status |
|---|---|---|---|---|---|
| 1 | 262,144 | 229.3 | 121.6 | 87 | observed |
Remaining fields
Identity, launch, related records, and measured speed are shown above. This is the rest of the normalized record.
- id
- qwen3-6-35b-a3b-ud-iq4-nl-rtx-3090-24gb-llama-cpp-tp1-sweep
- measured at
- 2026-05-13T18:15:37.846Z
- recipe id
- qwen3-6-35b-a3b-ud-iq4-nl-rtx-3090-24gb-llama-cpp-tp1
- schema version
- local-ai-registry/v1
metrics
- concurrency
- 1
- inference engine version
- b9102 (c8f8e2364, gg/spec-mtp-experiments)
- latest point at
- 2026-05-13T18:15:37.846Z
- max context tokens
- 262,144
- peak generation tps
- 121.6
- peak prompt tps
- 229.3
- point count
- 1
rows
| concurrency | context tokens | decode tok s | decode tok s per stream | output tokens | peak vram gb | prefill tok s | samples |
|---|---|---|---|---|---|---|---|
| 1 | 262,144 | 121.6 | 121.6 | 100 | 23.2 | 229.3 | 1 |
Provenance & metadata (1)
source
- kind
- leaderboard
- repository
- www.localmaxxing.com ↗