DraftFigures need sign-off before this week is published.
04b / Speed
glm-5.3-flash was quickest to start.
glm-5.3-flash had the quickest median first token of the 12 busiest models on the Requesty gateway in week 46 (Nov 9 to 15, 2026), 746 ms; for deepseek-v4-flash the slowest 1% of requests waited 36x its median.
A median hides the wait your slowest users feel. Look at p90 and p99 before you pick a model for anything interactive, and set timeouts from the tail, not the middle.
Let the gateway pick the fastest host →- Value
- 746 ms median time to first token
- Period
- Nov 9 to 15, 2026 (UTC)
- Compared with
- Nov 2 to 8, 2026
- Source
- Requesty gateway, shares, rates, ranks and latency only

Share
Post this chart
Preview on X1080 × 1080
>_
Requesty
@RequestyAI
Quickest median first token among the 12 busiest models on Requesty last week: glm-5.3-flash 746 ms, deepseek-v4.1-flash 834 ms, deepseek-v4-flash 1.14 s (Nov 9 to 15). requesty.ai/data/weekly/2026-w46/latency-spread

Platform
Image
Text
The link to this finding is added for you192 / 280
On desktop: copy the image, open X, paste. The text is filled in for you. On a phone, Share sends the image and text to the app in one go.
| Name | This week | Last week |
|---|---|---|
| glm-5.3-flash | 746 ms | n/a |
| deepseek-v4.1-flash | 834 ms | n/a |
| deepseek-v4-flash | 1.14 s | n/a |
| gpt-6-luna | 1.68 s | n/a |
| claude-haiku-5-5 | 1.75 s | n/a |
| nemotron-3-ultra-550b-a55b | 1.87 s | n/a |
| claude-sonnet-5 | 2.32 s | n/a |
| gpt-5.6-luna | 2.46 s | n/a |
| gpt-6.1-terra | 2.71 s | n/a |
| claude-sonnet-5-5 | 2.73 s | n/a |
More from week 46
New charts every Monday
Follow by RSS →For agents
- Markdown
- /data/weekly/2026-w46/data.md
- Latest
- /data/weekly/latest.json
- llms.txt
- /data/weekly/llms.txt
# Any week URL returns Markdown when asked for it
curl -H "Accept: text/markdown" \
https://www.requesty.ai/data/weekly/2026-w46Source: Requesty, The Gateway Charts, week 46 (Nov 9 to 15, 2026). CC BY 4.0.






