TokenMercato
live · updates every second
Mid
$0.060
0% under hosted
Spread
$0.0000
0 bps
Last
$0.030
4K tok
Capacity
194 tok/s
1 nodes

0.5B · Q4_K_M · 32.8K token context · per Mtok output

23:44:00high $0.030 · low $0.03004:31:00
1m candles · peak volume 16K tokens

Depth

Cumulative resting size by price

No resting depth on this side yet.

Recent fills

Coloured by aggressor side

PriceSizeTime (UTC)
$0.0304K02:50:01
$0.0304K02:50:01
$0.0304K02:50:01
$0.0304K02:50:01
$0.5004K22:32:51
$0.5004K22:00:19
$0.0304K21:50:57
$0.0304K21:50:52
$0.0304K21:49:51
$0.0304K21:48:31
$0.0304K21:46:54
$0.0304K21:46:38
$0.0304K21:45:26
$0.0304K21:44:44
$0.0304K21:44:23
$0.0304K21:44:01
$0.0304K21:43:55
$0.0304K21:43:39

Nodes quoting this model

1 online

NodeGPUAskTok/sCeiling
my-windows-nodeMeta Virtual Monitor$0.000194194

How this book matches

Rules applied in priority order

  1. 1
    Physical eligibility

    A node only sees the queue for models its VRAM and memory bandwidth can actually serve at its quoted speed. Everything below runs on that eligible set.

  2. 2
    Price priority

    The cheapest eligible ask fills first. Nothing outranks price — a better quote always jumps the line.

  3. 3
    Time breaks ties

    Equal asks fill oldest-first, so posting early is rewarded but never beats a lower price.

  4. 4
    Uptime is the last tie-break only

    Ranking by session length alone would hand every job to the single longest-lived node and starve new supply, so it settles ties and nothing more.

Book snapshot 2026-09-13 04:31:56 UTC.