OpenAI: gpt-oss-120b
by OpenAIOpenAI: gpt-oss-120b is a large language model from OpenAI. It costs $0.150 per million input tokens and $0.600 per million output tokens. Its context window is 131K tokens.
- Input / 1M tokens
- $0.150
- Output / 1M tokens
- $0.600
- Cached input / 1M
- $0.075
- Context window
- 131K
On repeated prefixes
tokens
Who serves it cheapest
23 hosts serve gpt-oss-120b. Same weights, same API - the price difference is pure margin and routing.
| Provider | Input / 1M | Output / 1M | Context | Throughput | Uptime 24h |
|---|---|---|---|---|---|
| Near AICheapestfp4 | $0.030 | $0.150 | 131K | - | 100.0% |
| Veniceint4 | $0.030 | $0.150 | 128K | - | 99.8% |
| CoreWeavefp4 | $0.030 | $0.170 | 131K | - | 100.0% |
| DekaLLMbf16 | $0.030 | $0.180 | 131K | - | 100.0% |
| AkashMLbf16 | $0.030 | $0.187 | 131K | - | 100.0% |
| DeepInfrabf16 | $0.037 | $0.170 | 131K | - | 98.2% |
| Crusoebf16 | $0.050 | $0.250 | 131K | - | 100.0% |
| Mancer 2fp8 | $0.050 | $0.250 | 131K | - | 99.6% |
| Novitafp4 | $0.050 | $0.250 | 131K | - | 99.7% |
| Sail Researchfp4 | $0.060 | $0.400 | 131K | - | 100.0% |
| DigitalOcean | $0.060 | $0.420 | 128K | - | 100.0% |
| $0.090 | $0.360 | 131K | - | 1.8% | |
| BaseTenfp4 | $0.100 | $0.500 | 128K | - | 100.0% |
| Amazon Bedrock | $0.150 | $0.600 | 131K | - | 100.0% |
| Nebiusfp4 | $0.150 | $0.600 | 131K | - | 100.0% |
| SiliconFlowfp8 | $0.150 | $0.600 | 131K | - | 89.4% |
| Phala | $0.150 | $0.600 | 131K | - | 99.5% |
| Together | $0.150 | $0.600 | 131K | - | 98.3% |
| Groq | $0.150 | $0.600 | 131K | - | 100.0% |
| Parasailfp4 | $0.100 | $0.750 | 131K | - | 100.0% |
| Mara | $0.150 | $0.750 | 131K | - | 96.4% |
| SambaNova | $0.140 | $0.950 | 131K | - | 100.0% |
| Cerebrasfp16 | $0.350 | $0.750 | 131K | - | 100.0% |
The spread between Near AI and Cerebras is 7.5× for identical weights. Quantization and context limits differ, so check both columns before switching.
Benchmarks
Independent scores published alongside the catalogue.
- Coding index
- 30.4
- Agentic index
- 3.7
About gpt-oss-120b
gpt-oss-120b is an open-weight, 117B-parameter Mixture-of-Experts (MoE) language model from OpenAI designed for high-reasoning, agentic, and general-purpose production use cases. It activates 5.1B parameters per forward pass and is optimized...
Specifications
| Model ID | openai/gpt-oss-120b |
|---|---|
| Provider | OpenAI |
| Context window | 131K tokens |
| Max output | 66K tokens |
| Input modalities | text |
| Output modalities | text |
| Knowledge cutoff | 2024-06-30 |
| Open weights | Yes - openai/gpt-oss-120b |
| Released | August 5, 2025 |
Cheaper alternatives
Models that cost less than gpt-oss-120b while keeping at least half its context window and every input modality it supports.
Frequently asked
How much does OpenAI: gpt-oss-120b cost?
$0.150 per million input tokens and $0.600 per million output tokens. Cached input reads cost $0.075 per million tokens.
What is the context window of OpenAI: gpt-oss-120b?
131K tokens, with up to 66K tokens of output per request.
Which provider serves OpenAI: gpt-oss-120b cheapest?
Near AI at $0.030 per million input tokens - 7.5× cheaper than Cerebras, the most expensive of the 23 hosts serving it.
Confirm against the source: OpenAI official pricing.