Compare/Step 3.7 Flash vs Llama 3.1 Nemotron Instruct 70B

Step 3.7 FlashvsLlama 3.1 Nemotron Instruct 70B

Side-by-side comparison of pricing, 12 benchmarks, and generation speed.

StepFun

Step 3.7 Flash

Input
$0.2/M
Output
$1.15/M
Speed
393 tok/s
TTFT
0.75s
NVIDIA

Llama 3.1 Nemotron Instruct 70B

Input
$1.2/M
Output
$1.2/M
Speed
303 tok/s
TTFT
4.03s

Winner by Category

Cheaper
Step 3.7 Flash
Faster (tok/s)
Step 3.7 Flash
Lower Latency
Step 3.7 Flash
Benchmarks (7-5)
Step 3.7 Flash

Pricing Comparison

MetricStep 3.7 FlashLlama 3.1 Nemotron Instruct 70B
Input ($/M tokens)$0.2$1.2
Output ($/M tokens)$1.15$1.2
Cost for 1M input + 100K output tokens:
Step 3.7 Flash$0.32
Llama 3.1 Nemotron Instruct 70B$1.32

Speed Comparison

Output Speed (tokens/s) — higher is better
Step 3.7 Flash
393 tok/s
Llama 3.1 Nemotron Instruct 70B
303 tok/s
Time to First Token (seconds) — lower is better
Step 3.7 Flash
0.75s
Llama 3.1 Nemotron Instruct 70B
4.03s

Benchmark Comparison

Data from Artificial Analysis API — 12 benchmarks

Intelligence Index
29.77.6
Coding Index
37.3
Math Index
11.0
GPQA Diamond
80.9%46.5%
MMLU-Pro
69.0%
LiveCodeBench
16.9%
AIME 2025
11.0%
MATH-500
73.3%
Humanity's Last Exam
19.9%4.6%
SciCode
40.0%23.3%
IFBench
67.3%30.7%
TerminalBench
35.6%4.5%
Step 3.7 Flash7 wins
5 winsLlama 3.1 Nemotron Instruct 70B

Frequently Asked Questions

Which is cheaper, Step 3.7 Flash or Llama 3.1 Nemotron Instruct 70B?

Step 3.7 Flash is cheaper overall. Its blended price (3:1 input/output ratio) is $0.44/M tokens vs $1.20/M for Llama 3.1 Nemotron Instruct 70B.

Which model performs better on benchmarks?

Step 3.7 Flash wins 7 out of 12 benchmarks compared to 5 for Llama 3.1 Nemotron Instruct 70B. See the detailed benchmark chart above for per-category results.

Which is faster for real-time applications?

Step 3.7 Flash generates tokens faster at 393 tok/s vs 303 tok/s. Step 3.7 Flash also has lower time-to-first-token (0.75s vs 4.03s).

When should I use Step 3.7 Flash vs Llama 3.1 Nemotron Instruct 70B?

Choose based on your priorities: Step 3.7 Flash for lower cost, Step 3.7 Flash for stronger benchmark performance, and Step 3.7 Flash for faster generation. For latency-sensitive apps, check the TTFT comparison above.