Mistralai Ministral 3 3B (2512)

model page · concurrency sweep 1–24 agents

Mistralai Ministral 3 3B (2512)

server model id: mistralai/ministral-3-3b-instruct-2512
non-thinkingKV Q4_0ctx 32k3B params

The official build vs the unsloth sibling at the same Q4_0 KV: 71 tok/s solo, smooth monotonic scaling to 404 tok/s at 22 agents — faster at every concurrency below the top end.

Key findings

  • Clean, all-content output: 0 reasoning tokens everywhere.
  • Much better behaved than the unsloth build at the same KV Q4_0: solo throughput is 71 tok/s vs 47 for unsloth/ministral-3-3b-instruct-2512 (same 4-bit KV), and the scaling curve is smooth and monotonic (71 → 400 tok/s) instead of unsloth's erratic 106–365 swings.
  • Peak 404 tok/s at 22 agents (5.7x); essentially flat at ~390–400 from 16–24. Near-identical absolute peak to the unsloth build (~410), so the two builds converge at high concurrency even though mistralai wins solo.
  • TTFT: 92ms solo → ~16.5s at 24; smooth growth, no outliers.
  • Per-agent speed: ~23–25 tok/s at high concurrency (vs ~174 solo burst metric).
  • Errors: 0 across all 24 runs.

Sweep — total concurrency 1–24

agentswall (s)contentreasontotalcomb tok/scomb allper-agentper-agent allscale %TTFT meanTTFT maxoktimeouterr
160.04,24804,2487171174.4174.4100%92ms92ms010
260.07,02607,026117117116.1116.1165%112ms125ms020
360.09,73009,73016216277.077.5228%3.5s3.5s030
460.012,100012,10020220272.572.6285%3.9s3.9s040
560.013,500013,50022522551.151.1317%6.0s6.1s050
660.014,961014,96124924948.448.4351%7.0s7.0s060
760.016,443016,44327427446.746.7386%7.6s7.6s070
860.017,130017,13028528541.841.8401%8.7s8.7s080
960.018,175018,17530330340.640.6427%9.2s9.2s090
1060.019,413019,41332332338.638.6455%9.7s9.7s0100
1160.019,277019,27732132138.438.4452%11.4s11.5s0110
1260.021,114021,11435235237.337.3496%11.6s11.7s0120
1360.021,976021,97636636636.136.1515%12.1s12.1s0130
1460.022,169022,16936936933.633.6520%12.6s12.6s0140
1560.023,092023,09238538532.632.6542%12.8s12.9s0150
1660.023,822023,82239739732.032.0559%13.4s13.5s0160
1760.122,402022,40237337328.728.7525%14.0s14.1s0170
1860.023,572023,57239339327.527.5554%12.4s12.5s0180
1960.023,481023,48139139126.526.5551%13.3s13.4s0190
2060.023,724023,72439539526.226.2556%14.7s14.8s0200
2160.023,393023,39339039025.725.7549%16.6s16.8s0210
22 peak60.024,238024,23840440424.824.8569%15.6s15.7s0220
2360.123,529023,52939239223.423.4552%16.3s16.5s0230
2460.124,026024,02640040023.123.1563%16.5s16.8s0240

* burst artifact — agents queued ~11–16s then generated in a short burst; the per-agent timer inflates the number. Trust combined throughput and TTFT columns. Peak row = highest combined (all-token) throughput.

Charts

Whole sweep at a glance — combined throughput, scaling efficiency, per-agent throughput and outcomes across every concurrency level.
Whole sweep at a glance — combined throughput, scaling efficiency, per-agent throughput and outcomes across every concurrency level.
Combined throughput (all tokens) vs concurrent agents. Annotated peak with the concurrency level where it occurred.
Combined throughput (all tokens) vs concurrent agents. Annotated peak with the concurrency level where it occurred.
Per-agent throughput erosion as slots contend. Red stars mark burst-timing artifacts (queued ≥ 10s then generated in a burst — inflated values, not real throughput).
Per-agent throughput erosion as slots contend. Red stars mark burst-timing artifacts (queued ≥ 10s then generated in a burst — inflated values, not real throughput).
Scaling efficiency relative to the single-agent baseline (100%).
Scaling efficiency relative to the single-agent baseline (100%).
Time to first token — mean (solid) and max (dashed). Log scale. Gaps mean no visible content ever arrived within the 60s window.
Time to first token — mean (solid) and max (dashed). Log scale. Gaps mean no visible content ever arrived within the 60s window.
Total tokens per 60s run — all tokens vs visible content only.
Total tokens per 60s run — all tokens vs visible content only.
How each run of N agents finished: completed, timed out (60s budget), or errored.
How each run of N agents finished: completed, timed out (60s budget), or errored.
The concurrency trade-off on one chart — combined throughput rises while per-agent speed falls.
The concurrency trade-off on one chart — combined throughput rises while per-agent speed falls.

Raw data & downloads