{
  "title": "Reasoning effort controls",
  "description": "Choose a documented reasoning depth, effort, or budget for a run.",
  "slug": "reasoning-effort-controls",
  "locale": "en",
  "seoTitle": "Reasoning effort controls compatibility — Can My Agent Use",
  "socialTitle": "Reasoning effort controls",
  "socialDescription": "Compare operator control over reasoning depth, effort, and budgets.",
  "llmSummary": "Reasoning effort controls let an operator choose a documented reasoning mode or budget; model selection alone does not prove this capability.",
  "audience": "Engineers balancing latency, cost, and task quality.",
  "contentKind": "feature",
  "status": "published",
  "tags": [
    "models",
    "reasoning",
    "latency",
    "cost"
  ],
  "updated": "2026-08-28T00:00:00.000Z",
  "published": "2026-08-28T00:00:00.000Z",
  "category": "models-context",
  "summary": "Choose a documented reasoning depth, effort, or budget for a run.",
  "specLabel": "Common product term",
  "aliases": [
    "thinking budget",
    "reasoning mode",
    "thinking level"
  ],
  "capabilityKind": "atomic",
  "parent": "models-and-context",
  "related": [
    "model-selection",
    "usage-metering"
  ],
  "relations": [],
  "highlight": false,
  "notes": [
    {
      "id": 1,
      "text": "Evidence checked 2026-08-28: ChatGPT Work on the web and ChatGPT desktop expose a reasoning control beneath the composer; Codex CLI exposes reasoning effort through /model, with documented Low through Ultra levels when supported."
    },
    {
      "id": 2,
      "text": "Evidence checked 2026-08-28: Claude web and desktop expose effort and extended-thinking controls; Claude Code documents per-session and configured effort levels, subject to model support."
    }
  ],
  "issues": [],
  "resources": [
    {
      "title": "Methodology",
      "href": "/methodology",
      "kind": "note"
    },
    {
      "id": "openai-models",
      "title": "OpenAI — Models in ChatGPT Work and Codex",
      "href": "https://learn.chatgpt.com/docs/models",
      "kind": "docs",
      "publisher": "OpenAI",
      "evidenceType": "documented",
      "reviewedAt": "2026-08-28"
    },
    {
      "id": "anthropic-change-model",
      "title": "Anthropic — Change model, effort, and thinking settings",
      "href": "https://support.claude.com/en/articles/8664678-change-the-model-effort-and-thinking-settings",
      "kind": "docs",
      "publisher": "Anthropic",
      "evidenceType": "documented",
      "reviewedAt": "2026-08-28"
    },
    {
      "id": "anthropic-code-model-config",
      "title": "Anthropic — Claude Code model configuration",
      "href": "https://code.claude.com/docs/en/model-config",
      "kind": "docs",
      "publisher": "Anthropic",
      "evidenceType": "documented",
      "reviewedAt": "2026-08-28"
    }
  ],
  "support": [
    {
      "harness": "chatgpt-web",
      "versions": [
        {
          "track": "current",
          "status": "yes",
          "noteIds": [
            1
          ],
          "target": {
            "kind": "hosted-observation",
            "revision": "2026-08-28 ChatGPT Work web documentation observation",
            "observedAt": "2026-08-28"
          },
          "environmentProfile": "hosted-default",
          "qualifiers": [
            {
              "type": "plan",
              "value": "available reasoning levels depend on the selected model and account eligibility"
            }
          ],
          "evidence": [
            {
              "resourceId": "openai-models",
              "type": "documented",
              "observedAt": "2026-08-28"
            }
          ]
        }
      ]
    },
    {
      "harness": "chatgpt-desktop",
      "versions": [
        {
          "track": "current",
          "status": "yes",
          "noteIds": [
            1
          ],
          "target": {
            "kind": "dated-documentation",
            "revision": "current ChatGPT desktop documentation",
            "observedAt": "2026-08-28"
          },
          "environmentProfile": "local-default",
          "qualifiers": [
            {
              "type": "plan",
              "value": "available reasoning levels depend on the selected model and account eligibility; Max may require enabling it in settings"
            }
          ],
          "evidence": [
            {
              "resourceId": "openai-models",
              "type": "documented",
              "observedAt": "2026-08-28"
            }
          ]
        }
      ]
    },
    {
      "harness": "codex-cli",
      "versions": [
        {
          "track": "current",
          "status": "yes",
          "noteIds": [
            1
          ],
          "target": {
            "kind": "dated-documentation",
            "revision": "current Codex CLI documentation",
            "observedAt": "2026-08-28"
          },
          "environmentProfile": "local-default",
          "qualifiers": [
            {
              "type": "runtime",
              "value": "documented selector includes Low, Medium, High, Extra high, Max, and Ultra when the selected model supports them"
            }
          ],
          "evidence": [
            {
              "resourceId": "openai-models",
              "type": "documented",
              "observedAt": "2026-08-28"
            }
          ]
        }
      ]
    },
    {
      "harness": "claude-web",
      "versions": [
        {
          "track": "current",
          "status": "yes",
          "noteIds": [
            2
          ],
          "target": {
            "kind": "hosted-observation",
            "revision": "2026-08-28 Claude web documentation observation",
            "observedAt": "2026-08-28"
          },
          "environmentProfile": "hosted-default",
          "qualifiers": [
            {
              "type": "runtime",
              "value": "effort and extended thinking are distinct controls and depend on model availability"
            }
          ],
          "evidence": [
            {
              "resourceId": "anthropic-change-model",
              "type": "documented",
              "observedAt": "2026-08-28"
            }
          ]
        }
      ]
    },
    {
      "harness": "claude-desktop",
      "versions": [
        {
          "track": "current",
          "status": "yes",
          "noteIds": [
            2
          ],
          "target": {
            "kind": "dated-documentation",
            "revision": "current Claude Desktop documentation",
            "observedAt": "2026-08-28"
          },
          "environmentProfile": "local-default",
          "qualifiers": [
            {
              "type": "runtime",
              "value": "effort and extended thinking are distinct controls and depend on model availability"
            }
          ],
          "evidence": [
            {
              "resourceId": "anthropic-change-model",
              "type": "documented",
              "observedAt": "2026-08-28"
            }
          ]
        }
      ]
    },
    {
      "harness": "claude-cli",
      "versions": [
        {
          "track": "current",
          "status": "yes",
          "noteIds": [
            2
          ],
          "target": {
            "kind": "dated-documentation",
            "revision": "current Claude Code documentation",
            "observedAt": "2026-08-28"
          },
          "environmentProfile": "local-default",
          "qualifiers": [
            {
              "type": "runtime",
              "value": "low, medium, high, xhigh, and max are documented where supported; ultrathink requests deeper reasoning for one turn"
            }
          ],
          "evidence": [
            {
              "resourceId": "anthropic-code-model-config",
              "type": "documented",
              "observedAt": "2026-08-28"
            }
          ]
        }
      ]
    }
  ],
  "html": "/features/reasoning-effort-controls",
  "markdown": "/features/reasoning-effort-controls.md",
  "json": "/api/v1/features/reasoning-effort-controls.json",
  "body": "This row covers an operator-visible control over reasoning effort, depth, or budget. It does not assert that hidden reasoning is exposed, and it does not treat ordinary model selection as an equivalent control.\n\nEvidence should name the available levels, defaults, model and plan restrictions, whether the setting changes output-token or tool-use budgets, and the scope at which it persists. Qualify controls that are only advisory or available through configuration rather than the primary interface."
}