Falcon3 1B Instruct

server model id: falcon3-1b-instruct

non-thinkingKV F16ctx 8k1B params

Fastest pure-content scorer — 839 tok/s at 13 agents, then rolls off as the 8k context saturates.

Key findings

Clean, all-content output: 0 reasoning tokens everywhere.
Strong scaling, with a peak and then a rolloff: combined throughput climbs from 187 tok/s solo to 839 tok/s at 13 agents (4.5x), holds ~730–840 through 18, then *declines* to 522 at 24 — unlike the other models, which plateaued. Under heavy contention with a small context/output cap, throughput rolls off.
Latency stays sane: TTFT 72ms–2.7s across the whole sweep (no 10s+ spikes like the thinking models).
Per-agent speed: falls from ~200 tok/s solo to ~24 at 24 agents.
Errors: 0 across all 24 runs.

Sweep — total concurrency 1–24

agents	wall (s)	content	total	comb tok/s	comb all	per-agent	per-agent all	scale %	TTFT mean	TTFT max	ok	timeout
1	60.0	11,214	11,214	187	187	202.5	206.8	100%	1.0s	1.0s	0	1
2	60.0	16,684	16,684	278	278	146.8	159.2	149%	72ms	91ms	0	2
3	60.0	20,889	20,889	348	348	126.6	135.8	186%	1.2s	1.2s	0	3
4	60.0	24,587	24,587	409	409	116.5	124.7	219%	1.5s	1.6s	0	4
5	60.0	28,624	28,624	477	477	101.1	106.2	255%	1.9s	1.9s	0	5
6	60.1	35,788	35,788	596	596	108.7	114.0	319%	748ms	785ms	0	6
7	60.1	38,712	38,712	645	645	99.1	105.1	345%	1.0s	1.0s	0	7
8	60.1	40,292	40,292	671	671	92.4	99.0	359%	1.0s	1.1s	0	8
9	60.1	44,005	44,005	733	733	88.3	97.6	392%	1.8s	1.9s	0	9
10	60.1	43,719	43,719	728	728	77.0	86.1	389%	1.3s	1.3s	0	10
11	60.1	46,032	46,032	766	766	78.4	89.0	410%	1.3s	1.4s	0	11
12	60.1	48,643	48,643	810	810	71.8	82.8	433%	2.0s	2.1s	0	12
13 peak	60.1	50,382	50,382	839	839	70.9	82.8	449%	677ms	704ms	1	12
14	60.1	47,925	47,925	798	798	61.7	73.4	427%	1.3s	1.4s	0	14
15	60.1	48,906	48,906	814	814	59.3	71.5	435%	1.9s	2.0s	0	15
16	60.1	49,630	49,630	826	826	57.4	70.6	442%	1.7s	1.7s	0	16
17	60.1	47,629	47,629	793	793	50.5	63.5	424%	2.7s	2.7s	0	17
18	60.1	48,878	48,878	814	814	48.8	62.8	435%	596ms	624ms	0	18
19	60.1	43,822	43,822	730	730	42.5	59.6	390%	2.0s	2.2s	0	19
20	60.1	42,649	42,649	710	710	38.7	56.9	380%	1.7s	1.7s	0	20
21	60.1	41,359	41,359	689	689	35.4	53.3	368%	1.5s	1.5s	0	21
22	60.1	37,553	37,553	625	625	30.5	50.8	334%	587ms	634ms	0	22
23	60.1	35,169	35,169	586	586	27.5	47.4	313%	827ms	860ms	0	23
24	60.1	31,331	31,331	522	522	23.6	43.6	279%	388ms	555ms	0	24

* 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.

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).

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.

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.

The concurrency trade-off on one chart — combined throughput rises while per-agent speed falls.

Raw data & downloads

report.md5 KB
sweep-summary.csv2 KB
sweep-summary.json11 KB
combined_throughput.png120 KB · original matplotlib export
combined_vs_per_agent.png105 KB · original matplotlib export
dashboard_1_24.png165 KB · original matplotlib export
outcome_breakdown.png57 KB · original matplotlib export
per_agent_throughput.png113 KB · original matplotlib export
scaling_efficiency.png76 KB · original matplotlib export
time_to_first_token.png57 KB · original matplotlib export
total_tokens_generated.png65 KB · original matplotlib export