gpt-oss-20b
gpt-oss-20b by OpenAI: 12 offers. From €0.031 per 1M input tokens, with EU processing from €0.04.
Calculate costs for this modelCompare offersView price history
Profile
| Profile | |
|---|---|
| Developer | OpenAI |
| Type | Language models |
| Origin | US |
| Input types | text |
| Context (tokens) | 131,000 |
| Max. output (tokens) | 117,964 |
| Reasoning | mit |
| Parameters (bn) | 21 |
| Open weights | ✓ yes |
| Licence | Apache-2.0 |
| VRAM at 4-bit (GB) | 16 (RTX PRO 4000) |
| Released | – |
Offers with processing in the EU (3)
| Provider | Country | Variant | Input | Output | Source | As of |
|---|---|---|---|---|---|---|
| cortecs | AT | standard | €0.04 | €0.15 | api.cortecs.ai | 14 Oct 2026 |
| OVHcloud | FR | standard | €0.04 | €0.15 | www.ovhcloud.com | 13 Oct 2026 |
| regolo.ai / Seeweb | IT | standard | €0.10 | €0.42 | api.regolo.ai | 13 Oct 2026 |
€ per 1M tokens. Source: providers' pricing pages, read and checked by searchit. As of 14 Oct 2026.
Offers outside the EU (9)
| Provider | Country | Variant | Input | Output | Source | As of |
|---|---|---|---|---|---|---|
| Google Vertex AI | US | batch | €0.031 | €0.11 | cloud.google.com | 14 Oct 2026 |
| AWS Bedrock | US | flex | €0.031 | €0.13 | pricing.us-east-1.amazonaws.com | 14 Oct 2026 |
| SiliconFlow | SG | standard | €0.036 | €0.16 | www.siliconflow.com | 2 Oct 2026 |
| Parasail | US | standard | €0.036 | €0.18 | www.parasail.io | 2 Oct 2026 |
| io.net | US | standard | €0.049 | €0.17 | api.intelligence.io.solutions | 14 Oct 2026 |
| Google Vertex AI | US | standard | €0.063 | €0.22 | cloud.google.com | 14 Oct 2026 |
| AWS Bedrock | US | standard | €0.063 | €0.27 | pricing.us-east-1.amazonaws.com | 14 Oct 2026 |
| Cloudflare Workers AI | US | standard | €0.18 | €0.27 | developers.cloudflare.com | 10 Oct 2026 |
| AWS Bedrock | US | priority | €0.11 | €0.47 | pricing.us-east-1.amazonaws.com | 14 Oct 2026 |
€ per 1M tokens. Source: providers' pricing pages, read and checked by searchit. As of 14 Oct 2026.
AI Act
Whether the developer has signed the Code of Practice for general-purpose AI models: unknown. What this means is explained under “EU and AI Act”. EU processing and AI Act
Benchmarks
| Benchmarks | |
|---|---|
| GPQA Diamond (Artificial Analysis) | 61.1 |
| Humanity's Last Exam (Artificial Analysis) | 5.3 |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 40.3 |
| Structured Output Benchmark (SOB), arXiv 2604.25359 | 0.86 |
| Structured Output Benchmark (SOB), arXiv 2604.25359 – json pass | 0.86 |
| Structured Output Benchmark (SOB), arXiv 2604.25359 – wertgenauigkeit | 0.69 |
| MMLU-Pro (TIGER-Lab Leaderboard) | 73.1 |