gpt-oss-20b is 37.0× cheaper per million tokens (blended). Pareto Latest has the larger context window (1,048,576 vs 131,072 tokens, 8.0×). gpt-oss-20b is open-weights; the other is not.
gpt-oss-20b vs Pareto Latest
Live catalog fields. Best-in-row is highlighted.
gpt-oss-20bopen
OpenAI · openai__gpt-oss-20b
Pareto Latest
~unbiased · ~unbiased__pareto-latest
Identity
| Field | openai__gpt-oss-20b | ~unbiased__pareto-latest |
|---|---|---|
| Provider | OpenAI | ~unbiased |
| Slug | openai__gpt-oss-20b | ~unbiased__pareto-latest |
| Status | live | live |
| Open weights | Yes | No |
| License | — | — |
| HuggingFace | openai/gpt-oss-20b | — |
Quality
| Field | openai__gpt-oss-20b | ~unbiased__pareto-latest |
|---|---|---|
| Quality index (0–100) | — | — |
Cost
| Field | openai__gpt-oss-20b | ~unbiased__pareto-latest |
|---|---|---|
| Input $ / 1M tokens | $0.02 | $0.80 |
| Cached input $ / 1M | $0.01 | $0.03 |
| Output $ / 1M tokens | $0.09 | $3.20 |
Context
| Field | openai__gpt-oss-20b | ~unbiased__pareto-latest |
|---|---|---|
| Max context tokens | 131,072 | 1,048,576 |
| Max output tokens | 32,768 | 131,072 |
Modalities
| Field | openai__gpt-oss-20b | ~unbiased__pareto-latest |
|---|---|---|
| Input | text | text, image |
| Output | text | text |