Microsoft: Phi 4 vs Llama 3.3 Nemotron Super 49B v1.5
Llama 3.3 Nemotron Super 49B v1.5 is cheaper than Microsoft: Phi 4 by about 100% on a blended (75% input / 25% output) basis. Microsoft: Phi 4 charges $0.070000 / $0.140000 per 1M input/output tokens; Llama 3.3 Nemotron Super 49B v1.5 charges $0.000000 / $0.000000.
| Field | Microsoft: Phi 4 | Llama 3.3 Nemotron Super 49B v1.5 |
|---|---|---|
| Input $/Mtok | $0.070000 | $0.000000 |
| Output $/Mtok | $0.140000 | $0.000000 |
| Cache read $/Mtok | – | – |
| Cache write $/Mtok | – | – |
| Input $/Mchar | n/a | 0.000000 |
| Output $/Mchar | n/a | 0.000000 |
| Context window | 16384 | 131072 |
| Tokenizer | – | llama3 |
Price history
Same workload, monthly cost
Illustrative monthly cost at 10M tokens/month, split per workload preset. Not a substitute for the calculator, which uses your own volume, cache and batch assumptions.
| Workload | Microsoft: Phi 4 | Llama 3.3 Nemotron Super 49B v1.5 |
|---|---|---|
| Coding agent | $0.74 | $0.00 |
| RAG chat | $0.81 | $0.00 |
| Summarisation (batch) | $0.77 | $0.00 |
| Classification | $0.71 | $0.00 |
| Chat assistant | $0.98 | $0.00 |
| Content generation | $1.26 | $0.00 |
Cite this page
Microsoft: Phi 4 vs Llama 3.3 Nemotron Super 49B v1.5: $0.070000/$0.140000 vs $0.000000/$0.000000 per 1M tokens, as of 2026-10-04 (Token Barometer, https://tokenbarometer.com/compare/microsoft_phi-4-vs-nvidia_nemotron-super-49b-v1).
Microsoft: Phi 4 page | Llama 3.3 Nemotron Super 49B v1.5 page | Open in calculator