| Claude Opus 5.5 | gpt-oss-20b | |
|---|---|---|
| Input, per 1M tokens | $4.00 | $0.018 |
| Output, per 1M tokens | $20.00 | $0.09 |
| Context window | 1M | 131K |
| Classification or routing (600 tokens in, 10 out), per 1,000 | $2.60 | $0.0117 |
| Chat reply (2,000 in, 400 out), per 1,000 | $16.00 | $0.072 |
| Retrieval-augmented answer (8,000 in, 500 out), per 1,000 | $42.00 | $0.189 |
| Confidence signal | no | yes |
| Tool calling | yes | yes |
| BANKING77, correct of 400 | 354 | 313 |
Which one should carry your traffic?
If gpt-oss-20b answers your requests as well as Claude Opus 5.5, moving that traffic saves 100% on chat-shaped requests. Averages cannot tell you whether it does for your requests. gpt-oss-20b exposes a confidence signal, so Gimbix can let it answer first and hand unsure cases to Claude Opus 5.5.
Claude Opus 5.5 profile · gpt-oss-20b profile · Measure it on your traffic