GPT OSS 120B
Open-weight GPT model for self-hosted reasoning and instruction-following workloads
Shogo Agent Index
Supported#86 of 9222.1range 15.7–28.5 · 32% of weight measured
Agentic coding
23
23% measured
Long-horizon autonomy
31
80% measured
Knowledge-work agents
17
32% measured
Computer use
12
estimated
Security & context learning
24
estimated
Agentic benchmarks
| Benchmark | Score | Normalised | Index weight | Harness / source |
|---|---|---|---|---|
| SWE-bench Verified | est. | 8% | predicted 58 | |
| Terminal-Bench 2.0 | 18.7% | 8% | Terminus 2 | |
| DeepSWE | est. | 6% | predicted 16 | |
| FrontierSWE | est. | 5% | predicted 2 | |
| CursorBench | est. | 4% | predicted 22 | |
| GSO-Bench | est. | 4% | predicted 4 | |
| METR Time Horizons | 56.6% | 10% | source | |
| Vending-Bench 2 | $-22 | 6% | source | |
| PostTrainBench | est. | 4% | predicted 14 | |
| APEX-Agents | 4.4% | 8% | source | |
| GDPval | est. | 7% | predicted 23 | |
| Remote Labor Index | est. | 5% | predicted 1 | |
| DeepResearch Bench | est. | 5% | predicted 43 | |
| OSWorld 2.0 | est. | 6% | predicted 1 | |
| OSWorld | est. | 4% | predicted 28 | |
| Cybench | est. | 5% | predicted 32 | |
| CL-bench | est. | 5% | predicted 15 |
Other benchmarks
| Benchmark | Category | Score | Source |
|---|---|---|---|
| Aider Polyglot | Coding | 41.8% | source |
| AlgoTune | Coding | 1.41× | source |
| CritPt | Reasoning | 1.1% | source |
| DTBench | Knowledge | 76.3% | source |
| Fiction.LiveBench | Long context | 44.4% | source |
| GPQA Diamond | Reasoning | 75.8% | source |
| LMCA | Knowledge | 22.1% | source |
| OTIS Mock AIME | Math | 88.9% | source |
| SciCode | Coding | 36.0% | source |
| SimpleBench | Reasoning | 22.1% | source |
API hosts
Live prices and uptime for every host serving GPT OSS 120B via OpenRouter.
| Host | Input | Output | Cache read | Agent $/M | Uptime 24h | Quant |
|---|---|---|---|---|---|---|
| Near AI | $0.030 | $0.15 | $0.015 | $0.020 | 99.9% | fp4 |
| DeepInfra | $0.037 | $0.17 | $0.018 | $0.030 | 38.5% | bf16 |
| AkashML | $0.030 | $0.19 | $0.030 | $0.040 | 100.0% | fp4 |
| CoreWeave | $0.030 | $0.17 | $0.030 | $0.040 | 98.9% | fp4 |
| DekaLLM | $0.030 | $0.18 | $0.030 | $0.040 | 99.2% | bf16 |
| DigitalOcean | $0.060 | $0.42 | $0.012 | $0.040 | 99.5% | — |
| Venice | $0.030 | $0.15 | $0.030 | $0.040 | 99.9% | int4 |
| Sail Research | $0.060 | $0.40 | $0.030 | $0.050 | 100.0% | fp4 |
| Crusoe | $0.050 | $0.25 | $0.050 | $0.060 | 100.0% | bf16 |
| Parasail | $0.10 | $0.75 | $0.055 | $0.090 | 100.0% | fp4 |
| $0.090 | $0.36 | — | $0.10 | 99.7% | — | |
| Mancer 2 | $0.085 | $0.38 | $0.085 | $0.10 | 96.4% | fp8 |
| Groq | $0.15 | $0.60 | $0.075 | $0.11 | 99.7% | — |
| SiliconFlow | $0.15 | $0.60 | $0.075 | $0.11 | 90.7% | fp8 |
| BaseTen | $0.10 | $0.50 | $0.10 | $0.12 | 99.9% | fp4 |
| BaseTen | $0.10 | $0.50 | $0.10 | $0.12 | 100.0% | fp4 |
| Nebius | $0.15 | $0.60 | $0.10 | $0.13 | 100.0% | fp4 |
| Amazon Bedrock | $0.15 | $0.60 | — | $0.17 | 100.0% | — |
| Amazon Bedrock | $0.15 | $0.60 | — | $0.17 | 100.0% | — |
| Together | $0.15 | $0.60 | — | $0.17 | 99.9% | — |
| Parasail | $0.15 | $0.75 | — | $0.18 | 100.0% | fp4 |
| SambaNova | $0.14 | $0.95 | — | $0.18 | 98.3% | — |
| PrimeIntellect | $0.35 | $0.75 | $0.32 | $0.34 | 98.3% | fp4 |
| Cerebras | $0.35 | $0.75 | $0.35 | $0.37 | 100.0% | fp16 |