GPT-5.6 Sol vs GPT-5.6 Terra on HealthBench Hard
updated August 16, 2026
A gap of 0.004 separates these two on the 1,000 hardest HealthBench conversations: GPT-5.6 Sol at 0.331, GPT-5.6 Terra at 0.327. The table adds what each one costs at list rates.
Side by side
 GPT-5.6 Sol | GPT-5.6 Terra | |
|---|---|---|
| score | 0.331 | 0.327 |
| rank | 2 of 9 | 3 of 9 |
| context window | 1.1M | 1.1M |
| price per 1M tokens, in / out | $5.00 / $30.00 | $2.00 / $12.00 |
| 1,000-exchange workload | $31.00 | $12.40 |
| released | 2026-07-09 | 2026-07-09 |
| license | proprietary | proprietary |
The workload row prices 1,000 exchanges of 2,000 input and 700 output tokens each, at the list rates current on August 16, 2026. GPT-5.6 models charge higher rates above 272K input tokens. GPT OSS models are open weights without vendor list pricing.
Reading the matchup
Inside the GPT-5.6 family the flagship premium buys almost nothing here: 0.004 on a 1,000-conversation set, well within re-run variation. Terra costs less than half as much on both input and output. Unless a workload already justifies Sol for other reasons, Terra is the same benchmark result at a lower bill.
Which scores higher on HealthBench Hard, GPT-5.6 Sol or GPT-5.6 Terra?
GPT-5.6 Sol. On the 1,000-conversation set it scores 0.331 to GPT-5.6 Terra's 0.327, a margin of 0.004 as of August 16, 2026.
Which is cheaper to run, GPT-5.6 Sol or GPT-5.6 Terra?
GPT-5.6 Terra. The same workload of 1,000 exchanges (2,000 input and 700 output tokens each) comes to $12.40 on it and $31.00 on the other.
Related comparisons
- 0.428 vs 0.331
- 0.331 vs 0.016
Each model's full page: GPT-5.6 Sol and GPT-5.6 Terra. Every other pairing lives in the score-difference matrix.