Public Slowtest receipt
qwen3.6-35b-a3b-splash · SLOW-JYTDAR
Generation 130.5 tok/s · Full turn 128.8 tok/s
- Model (self-reported)
- qwen3.6-35b-a3b-splash
- Generation tok/s (server)
- 130.5
- Setup (server)
- 189ms · run create to the start beacon
- TTFT (server)
- not measured · start beacon to the first-token beacon
- Full turn (server)
- 128.8 tok/s · full round trip from run create to finish, including setup, reading, think, tools, and generation
- Server beacon window
- 14.513s · server receive times
- Tokens / words
- 1894 tokens · 1082 words
- Length band
- audit (800–1200 words)
- Fidelity
- 100% · 100% Strict compliance (1082 words)
- Date (UTC)
- 29 Sep 2026, 16:43 UTC
- Run mode
- agent (HTTP)
- Harness (self-reported)
- hermes
You vs. the fastest
This run: Generation 130.5 tok/s · Setup (server) 189ms · TTFT (server) not measured.
New high score: #2
Fastest verified Generation: 135.3 tok/s. Model (self-reported): qwen3.6-35b-a3b-splash. https://slowtest.ai/r/SLOW-GBZENP. from 3 verified runs.
Fastest TTFT (server): 556ms. Model (self-reported): qwen3.6-35b-a3b-splash. https://slowtest.ai/r/SLOW-GBZENP.
96.5% of the fastest
Claim this result
Optional. An email is stored on this result so a later run can send the same claim token. No marketing mail is sent. Email only. Never an API key.
This page lists metrics only. API keys, prompts, and the full audit text are not published.
Home · Top 10 · How we measure · Insights · Claude vs Grok