Speed sweep
deepseek-v4-flash-ad-iq1-m-radeon-ai-pro-r9700-32gb-llama-cpp-tp3-sweep
deepseek-v4-flash-ad-iq1-m-radeon-ai-pro-r9700-32gb-llama-cpp-tp3-sweepRecord
- Measured
- 2026-08-04T22:09:38.332Z
- Points
- 1
Measured speed
| Concurrency | Context | Prefill | Decode | TTFT ms | Status |
|---|---|---|---|---|---|
| 1 | 782 | 274.3 | 30.9 | 1,917.7 | observed |
Remaining fields
Identity, launch, related records, and measured speed are shown above. This is the rest of the normalized record.
- id
- deepseek-v4-flash-ad-iq1-m-radeon-ai-pro-r9700-32gb-llama-cpp-tp3-sweep
- measured at
- 2026-08-04T22:09:38.332Z
- recipe id
- deepseek-v4-flash-ad-iq1-m-radeon-ai-pro-r9700-32gb-llama-cpp-tp3
- schema version
- local-ai-registry/v1
metrics
- concurrency
- 1
- latest point at
- 2026-08-04T22:09:38.332Z
- max context tokens
- 782
- peak generation tps
- 30.931
- peak prompt tps
- 274.291
- 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 | 782 | 30.931 | Unknown | 256 | 81.55 | 274.291 | 1 |
Provenance & metadata (1)
source
- kind
- leaderboard
- repository
- www.localmaxxing.com ↗