Model benchmarks
flash-next-swift-512k local LLM performance
As of November 2026, flash-next-swift-512k runs at up to 83.1 tok/s for local inference (best of 2 community benchmark runs across 1 GPU).
Model size
Unknown
Peak speed
83.1 tok/s
Average speed
80.0 tok/s
Avg PP
619.3 tok/s
Min memory
n/a
Max context
65,536 tokens
Avg output / run
60,684 tokens
Avg runtime / run
12m 10s
Avg quality
87.4
Benchmark runs
2
GPUs tested
1
Quality by task
Average LLM-judged quality (0–100) with the run-to-run spread shown as a P5–P95 band, overall and for each benchmark task, across all 2 runs. The low and high columns show how much the judge’s score varies between runs, and need at least two runs to display.
| Task | P5 (low) | Avg | P95 (high) |
|---|---|---|---|
| Overall | 85.8 | 87.4 | 88.9 |
| Agent Workflow | 87.2 | 89.1 | 91.1 |
| Code Generation | 77.0 | 78.9 | 80.8 |
| Role Play & Narrative | 91.4 | 93.4 | 95.5 |
| Research & Analysis | 87.5 | 87.9 | 88.4 |
Performance by hardware and tool
Every hardware/tool/quantization combination flash-next-swift-512k has been benchmarked on, ranked by peak token generation speed. Last updated November 2026.
| Hardware | Tool | Quant | Best tok/s | Avg tok/s | Memory | Context | Quality | Runs |
|---|---|---|---|---|---|---|---|---|
| AMD Radeon AI PRO R9700 | Strata | — | 83.1 tok/s | 80.0 tok/s | n/a | 65,536 tokens | 87.4 | 2 |
Benchmark runs
All 2 flash-next-swift-512k runs submitted so far — expand one for its hardware, quality breakdown and per-scenario detail.
Frequently asked questions
- Is flash-next-swift-512k good for coding?
- In our benchmarks, flash-next-swift-512k scores 78.9/100 for coding. It runs at about 80.0 tok/s, so if you want more speed, qwen3.8-27b-turbo is faster (~127.9 tok/s) and still scores well for coding (85.4/100).
- Is flash-next-swift-512k good for agentic (tool-using) tasks?
- In our benchmarks, flash-next-swift-512k scores 89.1/100 for agentic workflows. It runs at about 80.0 tok/s, so if you want more speed, Ornith-1.5-35B-A3B-BigBang-oQ8e-mtp is faster (~102.7 tok/s) and still scores well for agentic workflows (91.1/100).
- How fast is flash-next-swift-512k for local inference?
- Across 2 community benchmark runs, flash-next-swift-512k reaches up to 83.1 tok/s and averages 80.0 tok/s, with the fastest results on AMD Radeon AI PRO R9700.
- Which tools have been used to run flash-next-swift-512k?
- Benchmarks were submitted using Strata. Results are community-contributed and updated as new runs arrive.