gpt-oss-120b (high)
AI model by OpenAI. Real-time pricing and benchmark data.
Benchmarks
Editorial Analysis
Profile. gpt-oss-120b (high) is a budget-tier option from OpenAI at $0.59/M output tokens. Despite the low price, it scores competitively on Coding Index (30.4) and Intelligence Index (12.3) — making it a candidate for high-volume production workloads where unit cost dominates.
Latency. On latency, gpt-oss-120b (high) reports 230 tok/s output speed and 840ms time-to-first-token. Output speed is in the normal range — most chat and completion applications will be token-bound, not streaming-bound.
Cost model. For a workload of 1M input tokens + 100K output tokens, gpt-oss-120b (high) costs $0.21 per million input-equivalent requests. Multiply by your monthly request volume to project the bill, and compare against the speed/correctness profile above to decide if the trade-off is worth it for your workload.
Editorial summary generated from public pricing and benchmark data (Artificial Analysis API). All numbers are sourced — see the Artificial Analysis methodology for benchmark definitions. Last refreshed 2026-10-08.
Compare with similar models
| Model | Input | Output | Speed |
|---|---|---|---|
gpt-oss-120b (high)Current | $0.15 | $0.59 | 230 tok/s |
| Mistral Small 4 (Non-reasoning) | $0.15 | $0.60 | 155 tok/s |
| Mistral Small (Feb '24) | $0.15 | $0.60 | 161 tok/s |
| Mistral Small 4 (Reasoning) | $0.15 | $0.60 | 169 tok/s |
| NVIDIA Nemotron Nano 12B v2 VL (Reasoning) | $0.20 | $0.60 | 137 tok/s |
| Mistral Small (Sep '24) | $0.20 | $0.60 | 159 tok/s |