1stTop pick
Researched 19d agoGPT-6.1 Sol
- ARC-AGI-2
- 94.2%
- Output (from)
- $10.00 / 1M
Best Reasoning LLMs (2026)
As of 2026-11-14, LLM Reference's top pick for reasoning is GPT-6.1 Sol by OpenAI (ARC-AGI-2 94.2%, $10.00/1M output tokens), chosen from models that meet the podium eligibility rules. Second pick: Claude Fable 5.2 by Anthropic.
Last refreshed 2026-11-14. Next refresh: weekly.
Best reasoning and math LLMs in 2026, ranked by ARC-AGI-2 (ARC Prize), with GPQA Diamond for models without an ARC-AGI-2 score. Includes thinking-mode models and chain-of-thought specialists.
Verdict
Use GPT-6.1 Sol for reasoning today.
Claude Fable 5.2 is the runner-up, 0.5 points back on ARC-AGI-2.
2ndShortlist
Researched 5d agoClaude Fable 5.2
- ARC-AGI-2
- 93.75%
- Output (from)
- $40.00 / 1M
3rdShortlist
Researched 29d agoClaude Opus 5.5
- ARC-AGI-2
- 93.3%
- Output (from)
- $20.00 / 1M
How we rank
Reasoning models rank on ARC-AGI-2 (ARC Prize). Models without an ARC-AGI-2 score follow on GPQA Diamond. The two scores are never mixed.
- Eligibility — Reasoning flag or GPQA Diamond above the editorial floor used on this page.
- Primary ranking — ARC-AGI-2 (higher is better); models without an ARC-AGI-2 score follow by GPQA Diamond; newer release on ties.
- Variant collapse — We keep one row per model family (`familySlug` + parameter tier). When headline scores tie within ±0.5 pt (±10 Elo on Chatbot Arena), we pick the canonical SKU by lowest tracked input price, then GA over preview or limited access, then newest `release`. A folded sibling within the benchmark noise band can show a "Tied within margin" chip on that score cell.
- Podium eligibility — Preview, restricted, retired and retiring models never fill a podium slot. Preview means preview or invite-only access, or a model that is announced or pre-release. Restricted means restricted, partner, private or otherwise unavailable access. Retired means the model is past its retirement date or every provider route is retired. Retiring means the model, or any one of its provider routes, has a retirement date within the next 60 days. These models can still appear in the table below the podium.
- Pricing — Reasoning tiers are often priced separately — confirm provider SKUs.
| # | Model | Input $/1M | Output $/1M | |
|---|---|---|---|---|
| 1 | GPT-6 Astra RestrictedReasoningVisionTools ARC-AGI-2: 95% | $10.00 | $50.00 | |
| 2 | GPT-6.1 Sol ReasoningVisionTools ARC-AGI-2: 94.2% | $2.00 | $10.00 | |
| 3 | Claude Fable 5.2 ReasoningVisionTools ARC-AGI-2: 93.75% | $8.00 | $40.00 | |
| 4 | Claude Opus 5.5 ReasoningVisionTools ARC-AGI-2: 93.3% | $4.00 | $20.00 | |
| 5 | Gemini 4 Argon ReasoningVisionTools ARC-AGI-2: 91.25% | $2.00 | $10.00 | |
| 6 | Claude Fable 5.1 ReasoningVisionToolsVerify pricing ARC-AGI-2: 90% | $10.00 | $50.00 | |
| 7 | GPT-6 Sol ReasoningVisionTools ARC-AGI-2: 89.6% | $2.00 | $10.00 | |
| 8 | Gemini 3.8 Flash ReasoningVisionToolsVerify pricing ARC-AGI-2: 89.2% | $0.75 | $3.75 | |
| 9 | GPT-5.5 ReasoningVisionToolsVerify pricing ARC-AGI-2: 85% | $5.00 | $30.00 | |
| 10 | Gemini 3.7 Flash ReasoningVisionToolsVerify pricing ARC-AGI-2: 84.6% | $0.75 | $3.75 | |
| 11 | Claude Opus 4.7 ReasoningVisionTools ARC-AGI-2: 75.8% | $5.00 | $25.00 | |
| 12 | GPT-5.4 ReasoningVisionTools ARC-AGI-2: 73.3% | $2.50 | $15.00 | |
| 13 | Gemini 3.5 Flash ReasoningVisionToolsVerify pricing ARC-AGI-2: 72.1% | $1.50 | $9.00 | |
| 14 | Claude Opus 4.6 ReasoningVisionTools ARC-AGI-2: 68.8% | $5.00 | $25.00 | |
| 15 | Grok 4.6 ReasoningVisionTools ARC-AGI-2: 67.1% | $2.00 | $6.00 | |
| 16 | GLM-5.3-Flash ReasoningVisionTools ARC-AGI-2: 65.8% | $0.15 | $0.50 | |
| 17 | Gemini 3.6 Flash ReasoningVisionTools ARC-AGI-2: 60.4% | $0.75 | $3.75 | |
| 18 | Kimi K3 ReasoningVisionTools ARC-AGI-2: 60.4% | $3.00 | $15.00 | |
| 19 | GPT-6 Luna ReasoningVisionToolsVerify pricing ARC-AGI-2: 59.3% | $0.10 | $0.50 | |
| 20 | Claude Sonnet 4.6 ReasoningVisionTools ARC-AGI-2: 58.3% | $3.00 | $15.00 |
Honorable mentions
Next seats in this ranking. Lines below are from each model's stored description in LLMReference seed data—spot-check the model page before relying on a capability claim.
Gemini 4 Argon is a Google DeepMind reasoning model, announced on September 30, 2026 for trusted cyber defenders and testers in Google's Fairwind Program and made generally available on October 13, 2026 through the Gemini API, Gemini Enterprise Agent Platform, Google Antigravity and the Gemini app. Google built it for long-horizon software engineering, legal and finance work, cybersecurity defense and science. It takes text, image, video, audio and PDF input and returns text. Input and output share a 1,048,576-token limit; one synchronous request returns up to 65,536 output tokens, and long decodes extend output up to 1,048,576 tokens. It supports low, medium and high thinking levels, function calling, structured outputs, code execution, search grounding and implicit caching.
91.25%
ARC-AGI-2
Anthropic's generally available Mythos-class model for demanding reasoning, long-horizon agents, and ambitious coding, released September 1, 2026. Same underlying model as Claude Mythos 5.1 with cybersecurity and biology safeguards (flagged queries route to Opus 4.8 for cyber and Opus 5 for biology; those reroutes are not billed at Fable prices). Claude API id claude-fable-5-1. List price $10 per 1M input tokens and $50 per 1M output tokens; cache reads $0.25 per 1M (0.025x base input, 75% less than Fable 5's $1.00). 1M-token context, 128K max output, adaptive thinking always on (default effort high), text and image input to text output, vision, tool use. Available to Pro, Max, Team, and Enterprise users and on the Claude Platform natively, Amazon Bedrock, Google Cloud, Microsoft Foundry, and Claude Platform on AWS. Reliable knowledge cutoff June 2026.
90%
ARC-AGI-2
OpenAI GPT-6 Sol, announced September 22, 2026 with GPT-6 Luna, expands the GPT-6 family below flagship GPT-6 Astra for cost-efficient complex coding and agentic workflows. First-party OpenAI API id gpt-6-sol. Docs: 1,050,000 context, 128,000 max output, knowledge cutoff Apr 20, 2026, text and image input to text output, reasoning effort none/low/medium/high/xhigh/max, function calling, structured outputs, prompt caching, Batch and Flex at 50% of Standard, Fast mode at 2x, and computer use and code interpreter via Responses tools. Standard short-context list is $2 input / $0.20 cached input / $2.50 cache writes / $10 output per 1M tokens (50% below GPT-5.6 Sol promotional). Prompts with more than 272K input tokens use 2x input and cache rates and 1.5x output for the full request. Also generally available on Amazon Bedrock (openai.gpt-6-sol) and Microsoft Foundry. Available in ChatGPT Work and Codex for Plus, Pro, Business, Enterprise, and Edu.
89.6%
ARC-AGI-2
Compare Top Picks
Side-by-side comparison of the top picks by price, benchmark, and API access.
Browse Other Categories
Best LLMs for Code GenerationBest LLMs for RAGBest LLMs for Agents vs Agent Products (2026)Best LLMs for ClassificationBest Open Source LLMsBest Multimodal / Vision LLMsBest LLM for Translation in 2026Best AI Image Models in 2026Best AI Video Models in 2026Best Small Language Models (SLMs)Best LLMs for JSON / Tool useCheapest LLM APIs You Can Call Right NowBest Long Context LLMsBest Mainstream LLM APIs, RankedBest LLMs for EnterpriseBest Free LLMs You Can Use Right NowBest LLMs for WritingBest LLMs for MarketingBest LLMs for Customer Support
Frequently asked questions
What is the best LLM for reasoning?
LLM Reference's top pick for reasoning as of 2026-11-14 is GPT-6.1 Sol by OpenAI, with ARC-AGI-2 94.2%. Picks come from models that meet the podium eligibility rules, so the top pick may not lead the full table. Second pick: Claude Fable 5.2 by Anthropic (ARC-AGI-2 93.75%).
Which ranked LLM for reasoning has the lowest output price?
Among the ranked models shown on this page, GLM-5.3-Flash by Zhipu AI has the lowest listed output price: $0.500/1M output tokens, at rank 16.
What is the best open-weight LLM for reasoning?
GLM-5.3-Flash by Zhipu AI is the highest-ranked open-weight model shown on this page (rank 16, ARC-AGI-2 65.8%). License: MIT.
How is the reasoning ranking calculated?
Reasoning models rank on ARC-AGI-2 (ARC Prize). Models without an ARC-AGI-2 score follow on GPQA Diamond. The two scores are never mixed. Preview, restricted, retired and retiring models never fill a podium slot. Preview means preview or invite-only access, or a model that is announced or pre-release. Restricted means restricted, partner, private or otherwise unavailable access. Retired means the model is past its retirement date or every provider route is retired. Retiring means the model, or any one of its provider routes, has a retirement date within the next 60 days. These models can still appear in the table below the podium.
How often is this page updated?
Weekly. Last refreshed 2026-11-14.