NVIDIAPhi 4 Multimodal
Not yet ranked: too few independent results so far. Released 11 Dec 2024.
- Blended price
- $0.140/M
- $0.080 in · $0.320 out
- Output speed
- 17 tok/s
- First answer
- 0.84 s
- first token 0.84 s
- Context
- 128k
- Reasoning19
- Multimodal5
- Composite36
Benchmark results
One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | default | Source |
|---|---|---|
| GPQA Diamond (AA)not in index | 31.5%#485 | Artificial Analysis |
| Humanity's Last Exam (AA)not in index | 5.0%#377 | Artificial Analysis |
| Kagi LLM Benchmark | 17.1%#131 | Kagi LLM Benchmark |
Multimodal
| Benchmark | default | Source |
|---|---|---|
| MMMU-Pro | 14.5%#246 | Artificial Analysis |
Composite
| Benchmark | default | Source |
|---|---|---|
| AA Intelligence Index | 5.8#471 | Artificial Analysis |
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0001 | – | 18.3 s |
| Summarise a 30-page report | 12,000 / 600 | $0.0012 | – | 35.8 s |
| Code edit | 6,000 / 1,500 | $0.0010 | – | 1.5 min |
| Agentic coding session | 60,000 / 4,000 | $0.0061 | – | 3.9 min |
| Structured extraction | 2,000 / 200 | $0.0002 | – | 12.5 s |
See also
Data as of 9 Sept 2026. Compare with another model.