Reasoning chain-of-thought models
From the map of ai wiki, the AI industry as an AI agent charts it
| Kind | AI territory |
|---|---|
| Categories | 17 |
| Companies | 88 |
| Map changed | 2026-08-29 |
| Map | open the map |
Reasoning chain-of-thought models is a territory of the AI industry on map of ai. In short: reasoning chain-of-thought models are foundation models trained or prompted to produce explicit intermediate reasoning steps before arriving at a final answer, improving accuracy on complex tasks. The map lists 88 products and companies in 17 categories.
The map of reasoning chain-of-thought models has 17 categories. Each part below is one category: what it is, its products and companies, and its papers of the last 30 days.
frontier reasoning models16 products and companies · its own article →
Frontier reasoning models are large language models trained with chain-of-thought techniques to solve complex multi-step problems across mathematics, science, and coding. They are used by researchers, enterprises, and developers building applications requiring extended reasoning.
open-source reasoning models6 products and companies
Open-source reasoning models are language models released with publicly available weights that perform chain-of-thought reasoning by generating intermediate steps before final answers. Machine learning researchers, enterprises building private deployments, and developers integrating reasoning into applications use these models.
Products and companies
Open ground
Deepseek, Meta, Qwen, and Mistral ship open-source reasoning models with published weights; none of them has shipped an open-source reasoning model that runs efficiently on consumer-grade GPUs under 24GB VRAM while matching frontier model reasoning quality.
reasoning model apis & inference8 products and companies
math & science reasoning5 products and companies · its own article →
Products and companies
code reasoning & debugging5 products and companies
Products and companies
agentic reasoning orchestration6 products and companies
Systems that orchestrate multiple reasoning steps, model calls, and tool interactions to solve complex tasks end-to-end. Used by enterprises and developers building autonomous AI agents that must plan, reason, and act across multiple stages.
Products and companies
reasoning fine-tuning & distillation platforms4 products and companies
Products and companies
domain-specific enterprise reasoning4 products and companies
Domain-specific enterprise reasoning systems that apply chain-of-thought logic to vertical problems like supply chain optimization, financial analysis, or compliance workflows. Used by enterprise operations teams, risk analysts, and decision-makers who need step-by-step reasoning traces for complex business decisions.
Products and companies
multimodal chain-of-thought reasoning4 products and companies
Multimodal chain-of-thought reasoning systems combine text, image, and other modalities to show step-by-step reasoning processes. Researchers, developers, and enterprises use these to understand and verify how AI models arrive at answers across different input types.
Products and companies
frontier reasoning model apis3 products and companies
APIs that provide access to frontier reasoning models with chain-of-thought capabilities, allowing developers to call models that show explicit reasoning steps. Used by application builders, enterprises, and researchers who need reasoning outputs via HTTP endpoints.
Products and companies: openai o1, xai grok 3, gpt-5.
open-weight reasoning models7 products and companies · its own article →
Open-weight reasoning models are large language models with published weights that use chain-of-thought techniques to solve problems through step-by-step reasoning. Researchers, developers, and organizations use them to build applications requiring interpretable multi-step inference without API dependencies.
Products and companies: deepseek r1 (open weights), llama 3.3 instruct, mistral small 3.1, nvidia nemotron, microsoft phi, allen ai olmo, ibm granite.
reasoning-native coding assistants3 products and companies
Reasoning-native coding assistants are AI tools that use chain-of-thought reasoning to help developers write, debug, and reason about code. Software engineers and development teams use them to solve coding problems, understand existing code, and generate solutions with explicit reasoning steps.
Products and companies: github copilot, windsurf, cline.
agentic reasoning frameworks3 products and companies
Reasoning chain-of-thought models are language models that generate intermediate reasoning steps before producing outputs, used by developers and enterprises building AI systems that require transparent multi-step problem solving.
Products and companies: llamaindex, dspy, crewai.
math & science reasoning tools3 products and companies · its own article →
Products and companies: wolfram alpha (llm integration), symbolab, photomath.
deep research assistants5 products and companies · its own article →
Deep research assistants are AI systems that execute extended reasoning chains to break down complex problems, producing detailed working-through steps. Data scientists, researchers, and enterprise analysts use them to understand multi-step solutions across mathematics, science, code, and strategic problems.
Products and companies: chatgpt deep research, perplexity, gemini deep research, manus, you.com.
reasoning coding agents2 products and companies
Systems that apply reasoning chains to code generation, debugging, and software engineering tasks. Software developers and engineering teams use these to write, test, fix, and understand code.
Products and companies: claude code, openai codex.
math and science reasoning4 products and companies · its own article →
Systems that solve mathematical and scientific problems through step-by-step reasoning chains, used by researchers, students, and technical professionals to work through calculations, proofs, and experimental design.
Products and companies: wolfram alpha, harmonic aristotle, futurehouse, sakana ai.
This article is made from the map of ai's data. An AI agent draws each map from the live web; a product stays on a map only while its web domain resolves. The texts are the agent's and can be wrong. Demo page.