Hetzner Inference API — Performance

Name: Hetzner Inference API — Performance
Creator: robotwar.io

Measured from robotwar.io Hetzner OSINT · generated 2026-10-23T14:19:19Z · window: last 30 days (all-time totals in summary)

← Back to the Hetzner OSINT feed · see also the API status report

Calls in window

3 593

7 921 all time

Success rate

97.7%

3 510 completed

Latency p50

19.1 s

p95 73.1 s

Output tokens/s

59.8

median of completed calls

Rate limited

0.0%

1 responses

Server errors

0.6%

21 incl. gateway 504

Timeouts

0.0%

1 calls

Parse errors

0.0%

0 calls

Tokens in

17 243 795

32 682 721 all time

Tokens out

4 909 811

9 897 021 all time

Attempts per success

1.0

1.0 = never retried

Retries

calls beyond the first attempt

Collection schedule

All times on this page are Amsterdam (CET/CEST). A missed slot means no collection started within the tolerance of a scheduled time.

Collection runs at 00:00, 08:00, 16:00 Europe/Amsterdam · next expected 2026-10-24 00:00 CEST · arrived 88 of 90 expected since 2026-09-23 · 3 did not finish

Last completed collection started 2026-10-23 08:00 CEST and took 15 min.

Missed: 2026-09-29 00:00, 2026-09-29 08:00

started	duration	status
2026-10-23 16:00 CEST	—	running (unfinished)
2026-10-23 08:00 CEST	15 min	ok
2026-10-23 03:30 CEST	16 min	failed
2026-10-23 00:00 CEST	12 min	ok
2026-10-22 16:00 CEST	21 min	ok
2026-10-22 08:00 CEST	19 min	failed
2026-10-22 03:30 CEST	17 min	ok
2026-10-22 00:00 CEST	20 min	ok
2026-10-21 16:00 CEST	25 min	ok
2026-10-21 08:00 CEST	23 min	ok
2026-10-21 03:30 CEST	15 min	ok
2026-10-21 00:00 CEST	17 min	ok

Model selection

Chosen from measurement: a model is used once the endpoint offers it and it has answered reliably, inside the response-time limit.

step	model	how	since	why
corroborate	Qwen/Qwen3.6-35B-A3B	held	2026-10-23	Qwen3.8-27B leads by only 0.012 — under the 0.08 margin
deep	Qwen/Qwen3.6-35B-A3B	held	2026-10-23	Qwen/Qwen3.6-35B-A3B-FP8 leads by only 0.005 — under the 0.08 margin
lens	Qwen/Qwen3.6-35B-A3B	held	2026-10-23	Qwen3.8-27B leads by only 0.012 — under the 0.08 margin
synthesize	Qwen/Qwen3.6-35B-A3B	held	2026-10-23	Qwen/Qwen3.6-35B-A3B-FP8 leads by only 0.005 — under the 0.08 margin
tag	Qwen/Qwen3.6-35B-A3B	held	2026-10-23	Qwen/Qwen3.6-35B-A3B-FP8 leads by only 0.005 — under the 0.08 margin
trendwatch	Qwen/Qwen3.6-35B-A3B	held	2026-10-23	Qwen/Qwen3.6-35B-A3B-FP8 leads by only 0.005 — under the 0.08 margin

Calls by status, per day

Stacked by outcome, Amsterdam days.

Speed over time

Latency percentiles cover completed calls; the dashed line is the 7-day mean.

Latency per model

Median latency per Amsterdam day.

Qwen/Qwen3.6-35B-A3B-FP8

Qwen/Qwen3.6-35B-A3B

Per model

Window: last 30 days. First/last seen span the whole log.

model	calls	success	p50	p95	max	tok/s	avg out	429	5xx	timeouts	first seen	last seen
Qwen/Qwen3.6-35B-A3B-FP8	3 120	97.4%	20.2 s	75.4 s	286.9 s	55.9	1 322	1	21	0	2026-08-16	2026-10-23
Qwen/Qwen3.6-35B-A3B	473	99.8%	16.1 s	35.1 s	91.3 s	108.0	1 892	0	0	1	2026-10-18	2026-10-23

Model bench runs

Same prompt, same signals, every model.

when	model	parsed	latency	tokens in	tokens out
2026-08-16 19:26	Qwen/Qwen3.6-35B-A3B-FP8	yes	76.2 s	4 310	1 621
2026-08-16 19:25	Kimi-K2.7-Code	no	—	—	—
2026-08-16 19:17	GLM-5.2-NVFP4	yes	243.4 s	3 999	1 698
2026-08-16 18:57	DeepSeek-V4-Flash-0731	no	—	—	—

Per step

Repair calls are the single retry after an unparseable answer.

step	calls	success	p50	p95	avg in	avg out	repair calls
tag	1 172	99.0%	28.4 s	82.6 s	6 740	2 037	2
media	771	100.0%	629 ms	1142 ms	745	32	0
lens-cn	145	96.6%	19.2 s	61.3 s	4 886	1 366	17
lens-de	145	96.6%	17.7 s	60.5 s	4 806	1 319	19
lens-pl	141	96.5%	20.4 s	58.9 s	4 886	1 386	14
events	138	96.4%	22.9 s	59.6 s	10 687	1 702	13
lens-baltics	137	96.4%	23.9 s	91.0 s	5 254	1 752	1
trendwatch	134	90.3%	43.2 s	115.9 s	10 843	3 313	4
lens-ir	131	96.2%	22.6 s	52.0 s	4 627	1 449	4
lens-kp	130	96.2%	19.2 s	59.7 s	4 730	1 251	2
lens-nl	130	95.4%	15.7 s	41.2 s	3 129	1 073	3
lens-by	130	96.2%	18.3 s	43.4 s	3 136	1 103	4
synthesize	129	95.3%	27.6 s	99.6 s	5 781	2 145	0
lens-wider	129	95.3%	20.7 s	73.8 s	5 204	1 509	0
events-verify	21	100.0%	15.5 s	33.5 s	4 267	1 875	1
media-eval	10	100.0%	621 ms	1030 ms	746	32	0

Failures over time

Most recent failures

Last 20, newest first.

when	step	model	http	class	message
2026-10-19 14:10	trendwatch	Qwen/Qwen3.6-35B-A3B	—	timeout	Request timed out.
2026-10-06 22:22	trendwatch	Qwen/Qwen3.6-35B-A3B-FP8	504	gateway 504	504
2026-10-06 06:06	tag	Qwen/Qwen3.6-35B-A3B-FP8	504	gateway 504	504
2026-10-05 22:21	trendwatch	Qwen/Qwen3.6-35B-A3B-FP8	504	gateway 504	504
2026-10-05 22:14	synthesize	Qwen/Qwen3.6-35B-A3B-FP8	504	gateway 504	504
2026-10-05 07:06	lens-nl	Qwen/Qwen3.6-35B-A3B-FP8	504	gateway 504	504
2026-10-05 06:45	tag	Qwen/Qwen3.6-35B-A3B-FP8	504	gateway 504	504
2026-10-05 06:40	tag	Qwen/Qwen3.6-35B-A3B-FP8	504	gateway 504	504
2026-10-05 06:35	tag	Qwen/Qwen3.6-35B-A3B-FP8	504	gateway 504	504
2026-10-05 06:29	tag	Qwen/Qwen3.6-35B-A3B-FP8	504	gateway 504	504
2026-10-05 06:23	tag	Qwen/Qwen3.6-35B-A3B-FP8	504	gateway 504	504
2026-10-05 06:17	tag	Qwen/Qwen3.6-35B-A3B-FP8	504	gateway 504	504
2026-10-05 06:12	tag	Qwen/Qwen3.6-35B-A3B-FP8	504	gateway 504	504
2026-10-05 06:07	tag	Qwen/Qwen3.6-35B-A3B-FP8	504	gateway 504	504
2026-10-04 06:36	lens-wider	Qwen/Qwen3.6-35B-A3B-FP8	—	other	connection: Connection error.
2026-10-04 06:35	lens-wider	Qwen/Qwen3.6-35B-A3B-FP8	—	other	connection: Connection error.
2026-10-04 06:35	lens-wider	Qwen/Qwen3.6-35B-A3B-FP8	—	other	connection: Connection error.
2026-10-04 06:35	lens-wider	Qwen/Qwen3.6-35B-A3B-FP8	—	other	connection: Connection error.
2026-10-04 06:34	lens-wider	Qwen/Qwen3.6-35B-A3B-FP8	—	other	connection: Connection error.
2026-10-04 06:33	lens-by	Qwen/Qwen3.6-35B-A3B-FP8	—	other	connection: Connection error.

Time of day

Amsterdam hours across the whole window.

Median latency — darker is slower

Failure share — darker is worse

3593 calls placed across the window

Model availability over time

One availability check per day from /models; calls per model per Amsterdam day; last 14 days. Watch this when a model disappears.

The endpoint currently lists 4 models: Qwen/Qwen3.6-35B-A3B, Qwen/Qwen3.6-35B-A3B-FP8, Qwen/Qwen3.8-27B, Qwen3.8-27B. Not offered on the latest check: DeepSeek-V4-Flash-0731 (last seen 2026-08-16); GLM-5.2-NVFP4 (last seen 2026-08-16); Kimi-K2.7-Code (last seen 2026-08-16). This timeline shows if and when they return.

	10-11	10-12	10-13	10-14	10-15	10-16	10-17	10-18	10-19	10-20	10-21	10-22	10-23
DeepSeek-V4-Flash-0731	×	×	×	×	×	×	×	×	×	×	×	×	×
GLM-5.2-NVFP4	×	×	×	×	×	×	×	×	×	×	×	×	×
Kimi-K2.7-Code	×	×	×	×	×	×	×	×	×	×	×	×	×
Qwen/Qwen3.6-35B-A3B								23	71	96	96	93	93
Qwen/Qwen3.6-35B-A3B-FP8	93	94	97	209	193	149	176	157	43	67	64	94	48
Qwen/Qwen3.8-27B
Qwen3.8-27B

offered and used successfully (number = ok calls) offered, every call failed offered, not used not offered on /models no availability check that day

API surface over time

Capability flags are observed, never assumed.

taken	endpoint	sdk	models	change
2026-10-23 14:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-23 06:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-23 01:30	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-22 22:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-22 14:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-22 06:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-22 01:30	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-21 22:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-21 14:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-21 06:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-21 01:30	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-20 22:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-20 14:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-20 06:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-20 01:30	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-19 22:01	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-19 14:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-19 06:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-18 22:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-18 14:01	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-18 06:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-18 01:30	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-17 22:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-17 14:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-17 06:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-17 01:31	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-16 22:01	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-16 14:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-16 06:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-16 01:30	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-15 22:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-15 14:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-15 06:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-15 01:30	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-14 22:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-14 14:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-14 06:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-13 22:03	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-13 14:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-13 06:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-13 01:30	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-12 22:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-12 14:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-12 06:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-12 01:30	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-11 22:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-11 14:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-11 06:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-11 01:30	https://inference.hetzner.com/api/v1	2.49.0	4	no change
2026-10-10 22:00	https://inference.hetzner.com/api/v1	2.49.0	4	no change

Observed on the latest snapshot

thinking can be switched off: yesgateway 504 observed: norate-limit headers exposed: no

Models offered

Qwen/Qwen3.6-35B-A3BQwen/Qwen3.6-35B-A3B-FP8Qwen/Qwen3.8-27BQwen3.8-27B

Per-model first and last call

model	first call	last call	calls (all time)
DeepSeek-V4-Flash-0731	2026-08-16	2026-08-16	6
GLM-5.2-NVFP4	2026-08-16	2026-08-16	6
Kimi-K2.7-Code	2026-08-16	2026-08-16	10
Qwen/Qwen3.6-35B-A3B	2026-10-18	2026-10-23	473
Qwen/Qwen3.6-35B-A3B-FP8	2026-08-16	2026-10-23	7 426

Rate-limit posture

Client budget: 10 requests / 60 s · observed 429 responses in window: 1 (0.0% of calls)

header	latest value
retry-after	1