← Back to the Model Guide
Best model / capability
Reasoning: the best models
Models built for hard reasoning and planning.
How this page ranks
- Ranked by the Intelligence Index (Artificial Analysis, through OpenRouter), highest first. Ties go to the lower blended price.
- Reasoning is a capability, not a job: these are models tagged reasoning on their record.
- Each model must list text output on OpenRouter or on its own record.
- Prices and scores come from the market tables only, each with the day it is for. The price on a model's own record is not used.
- Missing data never ranks. A model with no confirmed score in the last 30 days or unknown inputs and outputs is listed apart, with the reason. Price is shown when one is confirmed, and is not needed to rank.
Market data as of 10 Nov 2026. Prices: OpenRouter models API (official). Score: Artificial Analysis, through OpenRouter (third-party).
Ranked: 35 of 180 models
Cheapest for reasoning →| # | Model | Intelligence Index | Score date | Blended / 1M | Price date |
|---|---|---|---|---|---|
| 1 |  Claude Fable 5.2Anthropic | 64.6 | 10 Nov 2026 | $16.00 | 10 Nov 2026 |
| 2 | Claude Opus 5.5Anthropic | 59.1 | 10 Nov 2026 | $8.00 | 10 Nov 2026 |
| 3 |  GPT-6 AstraOpenAI | 56.9 | 10 Nov 2026 | $20.00 | 10 Nov 2026 |
| 4 | GPT-6.1 SolOpenAI | 54.6 | 10 Nov 2026 | $4.00 | 10 Nov 2026 |
| 5 | Claude Sonnet 5.5Anthropic | 54.5 | 10 Nov 2026 | $4.00 | 10 Nov 2026 |
| 6 | GPT-6.1 TerraOpenAI | 51.3 | 10 Nov 2026 | $2.00 | 10 Nov 2026 |
| 7 |  Gemini 4 ArgonGoogle | 50.4 | 10 Nov 2026 | $4.00 | 10 Nov 2026 |
| 8 | Claude Fable 5.1Anthropic | 50.0 | 10 Nov 2026 | $20.00 | 10 Nov 2026 |
| 9 |  Kimi K3.1 ProMoonshotai | 46.5 | 10 Nov 2026 | $4.00 | 10 Nov 2026 |
| 10 |  Grok 4.5xAI | 38.8 | 26 Oct 2026 | $3.00 | 10 Nov 2026 |
| 11 |  Glm 5.2Cf | 33.7 | 26 Oct 2026 | $1.71 | 10 Nov 2026 |
| 12 | GLM 5.2Accounts | 33.7 | 26 Oct 2026 | $1.71 | 10 Nov 2026 |
| 13 | GLM-5.2Hf Zai Org | 33.7 | 26 Oct 2026 | $1.71 | 10 Nov 2026 |
| 14 | GLM-5.2Zai Org | 33.7 | 26 Oct 2026 | $1.71 | 10 Nov 2026 |
| 15 | Claude Opus 4.6Anthropic | 31.9 | 26 Oct 2026 | $10.00 | 10 Nov 2026 |
| 16 | Kimi K2 ThinkingMoonshotai | 22.0 | 26 Oct 2026 | $1.07 | 10 Nov 2026 |
| 17 | o3 ProOpenAI | 21.9 | 26 Oct 2026 | $35.00 | 10 Nov 2026 |
| 18 |  DeepSeek V3.2DeepSeek | 21.5 | 26 Oct 2026 | $0.290 | 10 Nov 2026 |
| 19 | Claude Sonnet 4.5Anthropic | 20.7 | 26 Oct 2026 | $6.00 | 10 Nov 2026 |
| 20 | o3OpenAI | 20.2 | 26 Oct 2026 | $3.50 | 10 Nov 2026 |
| 21 | o4 MiniOpenAI | 16.7 | 26 Oct 2026 | $0.963 | 10 Nov 2026 |
| 22 | Gemini 2.5 ProGoogle | 16.1 | 26 Oct 2026 | $3.44 | 10 Nov 2026 |
| 23 | o1OpenAI | 15.2 | 26 Oct 2026 | $26.25 | 26 Oct 2026 |
| 24 | R1 0528DeepSeek | 13.1 | 26 Oct 2026 | $0.912 | 10 Nov 2026 |
| 25 |  Qwen3 235B A22B Thinking 2507Qwen | 12.7 | 26 Oct 2026 | $1.21 | 10 Nov 2026 |
Showing the top 25 of 35 ranked models.
Not ranked, and why (145)
No confirmed score in the market tables in the last 30 days 138
- Arcee AI: Maestro Reasoning No confirmed Intelligence Index
- Claude 4.6 Opus Thinking No confirmed Intelligence Index
- Claude 4.6 Opus Thinking Low No confirmed Intelligence Index
- Claude 4.6 Opus Thinking Max No confirmed Intelligence Index
- Claude 4.6 Opus Thinking Medium No confirmed Intelligence Index
- Claude 4.7 Opus Thinking No confirmed Intelligence Index
- Claude Fable 5 (batch) No confirmed Intelligence Index
- Claude Fable 5.1 (batch) No confirmed Intelligence Index
- Claude Fable 5.2 (batch) No confirmed Intelligence Index
- Claude Opus 4 No confirmed Intelligence Index
- Claude Opus 4.7 (Fast) No confirmed Intelligence Index
- Claude Opus 4.8 (Fast) No confirmed Intelligence Index
- Claude Opus 4.8 Thinking No confirmed Intelligence Index
- Claude Opus 5.5 (batch) No confirmed Intelligence Index
- Claude Sonnet 4 No confirmed Intelligence Index
- Claude Sonnet 4.6 Thinking No confirmed Intelligence Index
- Claude Sonnet 5 (Free) No confirmed Intelligence Index
- Claude Sonnet 5.5 (batch) No confirmed Intelligence Index
- Command R+ (08-2024) No confirmed Intelligence Index
- DeepSeek R1 No confirmed Intelligence Index
- DeepSeek R1 (Turbo) No confirmed Intelligence Index
- DeepSeek R1 0528 Qwen3 8B No confirmed Intelligence Index
- DeepSeek R1 Distill Qwen 14B No confirmed Intelligence Index
- Deepseek R1 Distill Qwen 32B No confirmed Intelligence Index
- DeepSeek R1 Distill Qwen 32B No confirmed Intelligence Index
- DeepSeek R1 Llama 70B Abliterated No confirmed Intelligence Index
- DeepSeek R1 Qwen Abliterated No confirmed Intelligence Index
- DeepSeek Reasoner No confirmed Intelligence Index
- DeepSeek V3.1 Terminus (Thinking) No confirmed Intelligence Index
- DeepSeek V3.1 Thinking No confirmed Intelligence Index
- DeepSeek V3.2 Thinking No confirmed Intelligence Index
- DeepSeek V4 Flash (Thinking) No confirmed Intelligence Index
- DeepSeek V4 Pro (Thinking) No confirmed Intelligence Index
- DeepSeek V4 Pro Cheaper (Thinking) No confirmed Intelligence Index
- DeepSeek V4 Pro Thinking TEE No confirmed Intelligence Index
- DeepSeek/DeepSeek-V3.2-Exp-Thinking No confirmed Intelligence Index
- ERNIE 5.0 No confirmed Intelligence Index
- ERNIE-4.5-21B-A3B-Thinking No confirmed Intelligence Index
- ERNIE-4.5-VL-28B-A3B-Thinking No confirmed Intelligence Index
- Gemini 3 Flash Thinking No confirmed Intelligence Index
- Gemini 3.5 Flash Thinking No confirmed Intelligence Index
- Gemini 4 Argon (batch) No confirmed Intelligence Index
- Gemma 4 26B A4B Thinking No confirmed Intelligence Index
- Gemma 4 31B Thinking No confirmed Intelligence Index
- Gemma 4 31B Thinking TEE No confirmed Intelligence Index
- GLM 4.5 (Thinking) No confirmed Intelligence Index
- GLM 4.5 Air (Thinking) No confirmed Intelligence Index
- GLM 4.5V Thinking No confirmed Intelligence Index
- GLM 4.6 Thinking No confirmed Intelligence Index
- GLM 4.6 Turbo (Thinking) No confirmed Intelligence Index
- GLM 4.7 Flash Original Thinking No confirmed Intelligence Index
- GLM 4.7 Flash Thinking No confirmed Intelligence Index
- GLM 4.7 Original Thinking No confirmed Intelligence Index
- GLM 4.7 Thinking No confirmed Intelligence Index
- GLM 5 Original Thinking No confirmed Intelligence Index
- GLM 5 Thinking No confirmed Intelligence Index
- GLM 5.1 Thinking No confirmed Intelligence Index
- GLM 5.1 Thinking TEE No confirmed Intelligence Index
- GLM 5V Turbo Thinking No confirmed Intelligence Index
- GLM-5.2 No confirmed Intelligence Index
- GPT 5.1 Thinking No confirmed Intelligence Index
- GPT-4o No confirmed Intelligence Index
- GPT-6 Astra (batch) No confirmed Intelligence Index
- GPT-6.1 Sol (batch) No confirmed Intelligence Index
- GPT-6.1 Terra (batch) No confirmed Intelligence Index
- Grok 4.1 Fast Non-Reasoning No confirmed Intelligence Index
- Grok 4.1 Fast Reasoning No confirmed Intelligence Index
- Grok 4.2 Fast Non Reasoning No confirmed Intelligence Index
- Grok 4.20 (Reasoning) No confirmed Intelligence Index
- Grok 4.20 Beta Non-Reasoning No confirmed Intelligence Index
- Grok 4.20 Beta Reasoning No confirmed Intelligence Index
- Grok 4.20 Non-Reasoning No confirmed Intelligence Index
- Grok 4.20 Reasoning No confirmed Intelligence Index
- Grok-4-Fast-Non-Reasoning No confirmed Intelligence Index
- Grok-4-Fast-Reasoning No confirmed Intelligence Index
- Hermes 4 (Thinking) No confirmed Intelligence Index
- Hermes 4 Large (Thinking) No confirmed Intelligence Index
- Kimi K2 Thinking Original No confirmed Intelligence Index
- Kimi K2 Thinking Turbo No confirmed Intelligence Index
- Kimi K2 Thinking Turbo Original No confirmed Intelligence Index
- Kimi K2.5 Thinking No confirmed Intelligence Index
- Kimi K2.5 Thinking TEE No confirmed Intelligence Index
- Kimi K2.6 Thinking No confirmed Intelligence Index
- kimi-k2-thinking No confirmed Intelligence Index
- LFM2.5-1.2B-Thinking (free) No confirmed Intelligence Index
- Llama 3.3 70B Cu Mai No confirmed Intelligence Index
- LongCat Flash Thinking 2601 No confirmed Intelligence Index
- Mag Mell R1 No confirmed Intelligence Index
- Magistral Medium (latest) No confirmed Intelligence Index
- Magistral Medium (latest) No confirmed Intelligence Index
- Magistral Small No confirmed Intelligence Index
- Magistral Small 2506 No confirmed Intelligence Index
- MAI-DS-R1 No confirmed Intelligence Index
- MiMo V2 Flash (Thinking) No confirmed Intelligence Index
- MiMo V2 Flash (Thinking) Original No confirmed Intelligence Index
- MiniMax M3 Thinking No confirmed Intelligence Index
- Ministral 3 14B Reasoning 2512 No confirmed Intelligence Index
- Ministral 3 3B Reasoning 2512 No confirmed Intelligence Index
- Mistral Medium 3.5 Thinking No confirmed Intelligence Index
- Mistral Small 4 119B Thinking No confirmed Intelligence Index
- Nemotron 3 Nano Omni (free) No confirmed Intelligence Index
- Nemotron 3 Nano Omni 30B A3B Reasoning No confirmed Intelligence Index
- nemotron-content-safety-reasoning-4b No confirmed Intelligence Index
- NVIDIA Nemotron 3 Nano Omni No confirmed Intelligence Index
- Nvidia Nemotron 3 Super 120B Thinking No confirmed Intelligence Index
- o3 (batch) No confirmed Intelligence Index
- o3 Deep Research No confirmed Intelligence Index
- o3 Mini (batch) No confirmed Intelligence Index
- o4 Mini (batch) No confirmed Intelligence Index
- o4 Mini Deep Research No confirmed Intelligence Index
- o4 Mini High No confirmed Intelligence Index
- OpenAI o1-mini No confirmed Intelligence Index
- OpenAI o1-preview No confirmed Intelligence Index
- OpenAI o3-mini (Low) No confirmed Intelligence Index
- OpenAI o3-pro (2025-06-10) No confirmed Intelligence Index
- Phi-4-mini-reasoning No confirmed Intelligence Index
- Phi-4-Reasoning No confirmed Intelligence Index
- Qwen Plus 0728 (thinking) No confirmed Intelligence Index
- Qwen QwQ 32B Preview No confirmed Intelligence Index
- Qwen/Qwen3-VL-32B-Thinking No confirmed Intelligence Index
And 18 more.
What the model takes in and puts out is not known 7
- Qwen3 Max No input or output list from OpenRouter or the record
- Qwen3 Max Thinking No input or output list from OpenRouter or the record
- Qwen3 VL 235B A22B Thinking No input or output list from OpenRouter or the record
- Qwen3 VL 8B Thinking No input or output list from OpenRouter or the record
- Qwen3.8 Max No input or output list from OpenRouter or the record
- R1 No input or output list from OpenRouter or the record
- R1 Distill Llama 70B No input or output list from OpenRouter or the record
Automated from dated market data, not Enterprise DNA tested picks. Check the price and terms with the provider before you commit. Methodology: how the model numbers work.