GPT OSS 120B
by Charm Hyper · family gpt-oss · listed Apr 2026RTW
Open GPT reasoning model for self-hosted agents and controllable deployments
Current rates USD per 1M tokens
Input$0.176
Output$0.652
Cache read—
Cache write$0.088
Against the market
On the record
- Model id
gpt-oss-120b- Context window
- 131K tokens
- Max output
- 13K tokens
- Input modalities
- text
- Output modalities
- text
- Reasoning
- yes
- Tool calls
- yes
- Open weights
- yes
- Knowledge cutoff
- —
- Released
- 2026-04-13
- Last updated
- 2026-07-22
- Provider docs
- hyper.charm.land
Back of the envelope
≈
Also listed at
- Nvidiafree in · free out
- Kenarifree in · free out
- Privatemode AIfree in · free out
- Cortecsfree in · free out
- Weights & Biases$0.03 in · $0.17 out
- LLM Gateway$0.032 in · $0.14 out
- OpenRouter$0.037 in · $0.17 out
- Deep Infra$0.037 in · $0.17 out
- Kilo Gateway$0.039 in · $0.19 out
- Helicone$0.04 in · $0.16 out
- IO.NET$0.04 in · $0.40 out
- NanoGPT$0.05 in · $0.25 out
- SiliconFlow$0.05 in · $0.45 out
- NovitaAI$0.05 in · $0.25 out
- DInference$0.0675 in · $0.27 out