Same open weights, served by 3 providers. Weights: openai/gpt-oss-20b.
| Provider | Their model id | Serving | Input | Output | Cached in | Blended 3:1 | Source | |
|---|---|---|---|---|---|---|---|---|
DeepInfra | openai/gpt-oss-20b | flexFull precision0.8× standard | $0.024 | $0.112 | — | $0.046 | Verified 6h ago · Changed 8h ago | |
DeepInfra | openai/gpt-oss-20b | standardFull precision | $0.03 | $0.14 | — | $0.058 | Verified 6h ago · Changed 8h ago | |
DeepInfra | openai/gpt-oss-20b | priorityFull precision1.5× standard | $0.045 | $0.21 | — | $0.086 | Verified 6h ago · Changed 8h ago | |
AWS Bedrock | openai.gpt-oss-20b-1:0 | standardQuantization not disclosed | $0.07 | $0.30 | — | $0.128 | Verified 8h ago · Changed 8h ago | |
AWS Bedrock | openai.gpt-oss-20b | standardQuantization not disclosed | $0.07 | $0.30 | — | $0.128 | Verified 8h ago · Changed 8h ago | |
Groq | openai/gpt-oss-20b | standardQuantization not disclosed | $0.075 | $0.30 | $0.037 | $0.131 | Verified 8h ago · Changed 8h ago | |
AWS Bedrock | us-gov.openai.gpt-oss-20b-1:0 | standardQuantization not disclosed | $0.084 | $0.36 | — | $0.153 | Verified 8h ago · Changed 8h 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 | $5.8/mo |
| AWS Bedrock | $13/mo |
| AWS Bedrock | $13/mo |
| Groq | $13.5/mo |
| AWS Bedrock | $15.6/mo |
Cache hits use each provider's cached-input price where it has one. Cache write costs aren't included.