Reasoning effort with Meridian
Control how much compute a model spends thinking before it answers. Higher effort = deeper chains of thought, better accuracy on hard problems, and higher latency.
| Model | reasoning_effort | Notes |
|---|---|---|
| gpt-5-codex | lowmediumhigh | Default: medium |
| gpt-5-1 | lowmediumhigh | Default: medium |
| grok-4-20-reasoning | lowmediumhigh | Default: medium |
Important
Set max_tokens ≥ 2048 or use reasoning_effort=low to avoid empty responses. Reasoning tokens consume the max_tokens budget — if the model spends everything on internal thought, the visible answer is truncated to zero.