origin not confirmed
| Model | Coding index | tok/s | $/M in | $/M out | $ / session | Cache |
|---|---|---|---|---|---|---|
| Nemotron 3.5 Lightning | 26.8 | 287.6 | $0.10 | $0.25 | — | — |
| Nemotron 3 Ultra 550B | 49.3 | 132.0 | $0.60 | $3.60 | — | — |
| Nemotron 3 Super 120B | — | — | $0.08 | $0.40 | — | — |
| Nemotron 3 Nano 30B | 14.4 | 253.6 | $0.05 | $0.20 | — | — |
| Nemotron 3 Nano Omni 30B | 13.8 | 328.6 | $0.00 | $0.00 | — | — |
1 endpoints, cheapest first. This is the view the model-by-model tables cannot give you: what one provider actually offers, and at what precision.
| Model | Quant | $/M in | $/M out | Context | Uptime 30m |
|---|---|---|---|---|---|
| Nemotron 3 Nano Omni 30B | unknown | $0.00 | $0.00 | 256,000 | 99.2% |
Everyone serving these models, cheapest first, with the numeric precision each one runs. The lab is often not the cheapest place to buy its own model.
| Model | Served by | Quant | $/M in | $/M out | Uptime 30m |
|---|---|---|---|---|---|
| Nemotron 3.5 Lightning | DeepInfra | bf16 | $0.08 | $0.20 | 99.6% |
| Nemotron 3.5 Lightning | CoreWeave | bf16 | $0.10 | $0.25 | 100.0% |
| Nemotron 3.5 Lightning | Venice | fp4 | $0.10 | $0.25 | 100.0% |
| Nemotron 3 Ultra 550B | DeepInfra | fp4 | $0.50 | $2.20 | 91.2% |
| Nemotron 3 Ultra 550B | BaseTen | fp4 | $0.60 | $2.40 | 98.8% |
| Nemotron 3 Ultra 550B | Venice | fp8 | $0.62 | $3.12 | 87.4% |
| Nemotron 3 Ultra 550B | Together | unknown | $0.60 | $3.60 | 98.2% |
| Nemotron 3 Super 120B | DeepInfra | bf16 | $0.08 | $0.40 | 92.8% |
| Nemotron 3 Super 120B | DigitalOcean | unknown | $0.17 | $0.36 | 92.9% |
| Nemotron 3 Super 120B | Nebius | fp4 | $0.30 | $0.90 | 98.0% |
| Nemotron 3 Nano 30B | Crusoe | fp8 | $0.05 | $0.20 | 98.2% |
| Nemotron 3 Nano 30B | DeepInfra | fp4 | $0.05 | $0.20 | 96.3% |
| Nemotron 3 Nano 30B | Novita | fp4 | $0.05 | $0.20 | 0.0% |
| Nemotron 3 Nano Omni 30B | Nvidia | unknown | $0.00 | $0.00 | 99.2% |