Concept card

Reasoning Models — Claude Extended Thinking, GPT-o4, Gemini Thinking

As of 2026-07-23

Edited by Cédric-Olivier Mâry, sole editor and publication director — 27 years in the SAP ecosystem. How these pages are produced, and where automation is used.

What is Reasoning Models — Claude Extended Thinking, GPT-o4, Gemini Thinking?

Reasoning models spend 10-100× more inference compute per hard query via an internal chain-of-thought, and the right architecture routes only the 5-15% of genuinely hard agent invocations to them, not every call.

What it is

Reasoning models are the 2026 architectural inflection that the Stanford HAI AI Index 2026 documents as the technical underpinning of the agentic-AI category. They differ from baseline LLMs in one crucial axis: at inference time they spend additional compute on an internal chain-of-thought before producing the final answer — the model's 'thinking' is itself a computational step, not just a prompt-engineering trick. The three production-grade variants in 2026 are Anthropic's Claude Extended Thinking, OpenAI's GPT-o4, and Google's Gemini Thinking.

Why it matters

The benchmark signature is unmistakable: GPQA Diamond jumps from 40-50% baseline to 70-85% with reasoning enabled, and reasoning models cross competition-math thresholds baseline models never reach.
The cost is real — 30-120 seconds per hard answer versus 2-5 seconds for a baseline LLM, and 5-20× higher token cost per query.
Each vendor ships a distinct variant into a distinct product: Claude Extended Thinking (visible traces, SAP's autonomous supply-chain agents), GPT-o4 (hidden by default, Microsoft Copilot Studio), Gemini Thinking (visible, Databricks Mosaic AI).

Key points

Reasoning models spend variable inference-time compute on internal chain-of-thought before answering — the 2026 architectural inflection per Stanford HAI AI Index 2026.
Benchmark signature on GPQA Diamond — 70-85% reasoning vs 40-50% baseline; the gap is the canonical 'reasoning mode value' metric.
Three production variants — Claude Extended Thinking (Anthropic, visible traces, used by SAP for supply-chain agents) · GPT-o4 (OpenAI, hidden by default, Copilot Studio) · Gemini Thinking (Google, Vertex AI, Databricks Mosaic AI).
Trade-off — 30-120s latency vs 2-5s baseline · 5-20× per-query cost · only justified for the 5-15% of agent invocations where the question is genuinely hard.
Hybrid routing pattern — Joule Studio 2.0 (June 2026) routes automatically between reasoning and baseline; the architecture decision a consultant defends with the customer's CFO.
Reasoning Models — Claude Extended Thinking, GPT-o4, Gemini Thinking is mastered only when it changes a named buyer decision.
Start with the semantic contract and control model before demonstrating the tool.
Use current SAP, analyst, study, KG, and news signals as evidence, not decoration.
Separate verified facts from directional trends and modeled assumptions.
Define owner, metric, threshold, support path, and rollback before scaling.

Terms used on this page

Extended thinking: Anthropic's term for the variable-inference-time-compute mode of Claude where the model generates a visible chain-of-thought trace before producing the final answer; the thinking-budget parameter is controllable by the calling agent.
GPT-o4: OpenAI's 2026 reasoning-model variant; hidden chain-of-thought by default, opt-in summarised traces; the variant surfaced by Microsoft Copilot Studio for high-complexity agent workflows.
Gemini Thinking: Google's 2026 reasoning-model variant; visible thinking, native Vertex AI integration; the variant most embedded in Databricks Mosaic AI agents.
Hybrid routing: The architectural pattern where an agent platform routes each invocation between a reasoning model (5-15% of hard questions) and a baseline LLM (85% routine) to balance accuracy vs latency vs cost; Joule Studio 2.0 (June 2026) implements this natively.
Decision owner: The accountable person who accepts the trade-off and funds the next action.
Semantic contract: The shared definition of business terms, metrics, entities, and access rules used by tools and teams.
Control plane: The layer that applies policy, access, lineage, monitoring, and escalation across the operating model.
Evidence grade: A label that separates verified fact, directional signal, modeled assumption, and field observation.

Sources

Full card available to members. What the full card adds: the full decision framework · the SAP vs Snowflake / Databricks / Fabric comparison · the common pitfalls and their fix · the cheat sheet · the architecture schemas · the code blocks.

Open in the app →