Models /gpt-oss-safeguard-20b /Where to run
Provider guide

Where to run gpt-oss-safeguard-20b

14 live offers tracked — output prices vary 2.1× between the cheapest and most expensive host, so the provider choice matters as much as the model choice.

CHEAPEST
$0.040 / $0.15
per 1M tokens in / out
FASTEST MEASURED
373 tok/s
measured throughput · $0.075 input / $0.30 output per million tokens
DeepInfra$0.030 in/1M$0.14 out/1M131Kbfloat166h agoNovita AICHEAPEST$0.040 in/1M$0.15 out/1M131K6h agoOpenRouter$0.075 in/1M$0.30 out/1M131K17m agoCoreWeavezero-retention$0.030 in/1M$0.13 out/1M94 tok/s131Kfp415m agoDeepInfrazero-retention$0.030 in/1M$0.14 out/1M97 tok/s131Kbf1615m agoAmazon Bedrockzero-retention$0.070 in/1M$0.15 out/1M323 tok/s131K15m agoNovita AIzero-retention$0.040 in/1M$0.15 out/1M120 tok/s131Kfp415m agoPhalazero-retention$0.040 in/1M$0.15 out/1M57 tok/s131K15m agoAmazon Bedrockzero-retention$0.070 in/1M$0.15 out/1M71 tok/s131K15m agoSiliconFlowzero-retention$0.040 in/1M$0.18 out/1M50 tok/s131Kfp86h agoTogether AIzero-retention$0.050 in/1M$0.20 out/1M82 tok/s131K15m agoGoogle Vertex AIzero-retention$0.070 in/1M$0.25 out/1M127 tok/s131K15m agoFireworks AIzero-retention$0.070 in/1M$0.30 out/1M89 tok/s131K15m agoGroqzero-retention$0.075 in/1M$0.30 out/1M373 tok/s131K15m ago
Full specs, hardware verdicts and benchmarks on thegpt-oss-safeguard-20b model page →