Same open weights, served by 2 providers.
| Provider | Their model id | Serving | Input | Output | Cached in | Blended 3:1 | Source | |
|---|---|---|---|---|---|---|---|---|
DeepInfra Deprecated, 2026-10-08 | nvidia/NVIDIA-Nemotron-Nano-9B-v2 | standardFull precision | $0.04 | $0.16 | — | $0.07 | Verified 5h ago · Changed 8h ago | |
AWS Bedrock | nvidia.nemotron-nano-9b-v2 | standardQuantization not disclosed | $0.06 | $0.23 | — | $0.103 | Verified 7h ago · Changed 7h ago |
Same weights don't always mean identical outputs: quantization, context limits and serving stacks can change quality, especially for tool calling. Prices are USD per 1M tokens; hyperscaler prices are for the reference region shown.
| DeepInfra bfloat16 | $7.2/mo |
| AWS Bedrock | $10.6/mo |
Cache hits use each provider's cached-input price where it has one. Cache write costs aren't included.