Llama 4 Maverick vs Nemotron 70B Instruct
Side-by-side comparison of API pricing, specs, benchmarks, and capabilities
|
Dby DeepInfra
|
Nby NVIDIA
|
|
|---|---|---|
| Overview | ||
| Status | Current open weights Open weights | Current open weights Open weights |
| Released | Apr 6, 2025 | Jun 1, 2025 |
| Pricing per million tokens | ||
| Input | $0.150/Mtok | $1.20/Mtok |
| Output | $0.600/Mtok | $1.20/Mtok |
| Blended avg | $0.375/Mtok | $1.20/Mtok |
| Specifications | ||
| Context window | 1M tokens | 128K tokens |
| Parameters | 17B (128 experts) | 70B |
| Speed (TPS) | — | — |
| Modalities | ||
| Input |
textimage
|
text
|
| Benchmarks sources: Vellum LLM Leaderboard / HuggingFace Open LLM Leaderboard | ||
| Avg benchmark score | 65.1 | — |
| Perf / dollar | 173.6 | — |
| GPQA Diamond | 69.8 | — |
| LiveCodeBench | 41 | — |
| MMMLU | 84.6 | — |
| Independent composite scores each on its own scale — not part of the average above | ||
| HF Open LLM v2 avg (0–100) | — | 36.9 |
| Providers | ||
| Available from |
DeepInfra — $0.150/$0.600/Mtok
|
NVIDIA — $1.20/$1.20/Mtok
|
Cost at scale — 1M tokens (50/50 input/output)
| Volume | Llama 4 Maverick | Nemotron 70B Instruct | Savings |
|---|---|---|---|
| 1M tokens | $0.38 | $1.2 | $0.83 (69.2%) |
| 10M tokens | $3.75 | $12 | $8.25 (68.8%) |
| 100M tokens | $37.5 | $120 | $82.5 (68.8%) |
| 1000M tokens | $375 | $1200 | $825 (68.8%) |
Summary
Llama 4 Maverick by DeepInfra costs $0.150/Mtok input and $0.600/Mtok output, with a 1M-token context window. It supports text, image input.
Nemotron 70B Instruct by NVIDIA costs $1.20/Mtok input and $1.20/Mtok output, with a 128K-token context window. It supports text input.
On a blended cost basis, Llama 4 Maverick is 68.8% cheaper than Nemotron 70B Instruct. It also has a larger context window.
The two aren't directly comparable on average benchmark score: Llama 4 Maverick has published per-benchmark results, while Nemotron 70B Instruct does not yet — it is measured today on independent composites (see the table above).
Note: Pricing is per million tokens. Actual costs vary with usage patterns, prompt caching, and batch discounts. Always verify against official provider pricing pages.