Skip to main content

Includes 694 hourly rounds up to 2026-11-16 from NIMStats by Mauro Druwel (source), measured with a different prompt and timeout. Our own probe adds a round every 30 minutes.

nvidia

nemotron-3-super-120b-a12b

Unstable

Rank 15 of 16 Last probed (25 h ago)

nvidia/nemotron-3-super-120b-a12b answered 5.8% of probes on the NVIDIA NIM free API in the last 30 days, with a median time to first token of 1.68 s and 44.7 tok/s. It ranks 15 of 16 tracked models.

5.8% Uptime 40/695 probes
28/ 100 Score 9 round wins
1.68 s First token p95 5.28 s
10.8 s Full reply median
3.60 s Best reply fastest in range
44.7tok/s Throughput median

Model vs. median

Against the median of models that answered, last 30 days.

Uptime
5.8% 77.3%
72 pts below median
First token
1.68 s 16.6 s
9.9× better than median
Full reply
10.8 s 22.6 s
2.1× better than median
Throughput
44.7 tok/s 18.0 tok/s
2.5× better than median

nemotron-3-super-120b-a12b median of answering models

Error breakdown

655 failed of 695 probes, last 30 days.

FailureProbesShare of failures
Not served (404)655 100%

Time to first token

Each probe. Lower is better.

No probe was answered in this range, so there is no latency to plot.

Throughput

Tokens per second after the first token, each probe. Higher is better.

No answered probe in this range.

Availability heatmap

Last 7 days by UTC hour.

all answered some failed none answered not probed

Run history

Last 20 probes.

TimeResultFirst tokenFull replySpeed
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a
Not served (404)n/an/an/a

Call nemotron-3-super-120b-a12b

The endpoint is OpenAI-compatible.

Create a free key at build.nvidia.com, then:

curl https://integrate.api.nvidia.com/v1/chat/completions \
  -H "Authorization: Bearer $NVIDIA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "nvidia/nemotron-3-super-120b-a12b",
    "messages": [{"role": "user", "content": "Hello"}],
    "max_tokens": 256,
    "stream": true
  }'

Other models answering

Top of the leaderboard right now.