OpenAI
gpt-oss-120b
Released Aug 5, 2025 · 178B tokens this week · 5 providers
Overview
An open-weight mixture-of-experts model from OpenAI with configurable reasoning effort. Because the weights are public it is served by many accelerator vendors at a fraction of API-model pricing. A strong default when you want GPT-family behaviour on commodity inference.
Providers
Reference data — not yet measured from live traffic
| PROVIDER | MAX OUT | OUTPUT /M | QUANT | |||||
|---|---|---|---|---|---|---|---|---|
DeepInfraCheapest | 131K | 33K | $0.090 | $0.45 | 620ms | 130 | 98.92% | fp8 |
Novita AI | 131K | 33K | $0.095 | $0.48 | 700ms | 110 | 98.61% | fp8 |
Fireworks AI | 131K | 33K | $0.10 | $0.50 | 420ms | 240 | 99.58% | Full |
Groq | 131K | 33K | $0.10 | $0.52 | 209ms | 780 | 99.86% | Full |
Cerebras | 66K | 33K | $0.11 | $0.53 | 190ms | 1450 | 99.71% | Full |
Latency p50 reflects a provider’s API endpoint responsiveness — the round-trip to its API, not per-token inference time. These figures, with throughput and uptime, are reference data until the gateway aggregates its own traffic.
Price comparison
Blended $ per 1M tokens
Blended rate per 1M tokens, weighted one part prompt to three parts completion — roughly the shape of a chat workload. Your mix will move the number.
Call it
OpenAI-compatible — swap the base URL and go
curl https://model.cards/api/v1/chat/completions \
-H "Authorization: Bearer $MODELCARDS_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "openai/gpt-oss-120b",
"messages": [
{ "role": "user", "content": "Summarize the tradeoffs of speculative decoding." }
]
}'