Early build · scores computed by the deterministic engine from real, sourced developments (30 days) · weights v1.0-prior
Latest › Models › Compare › For reasoning
Vikshy Compare

AI models built for reasoning

AI models built for step-by-step reasoning, the ones that think through hard problems before answering, ordered by tier then price. 24 in the catalogue.

ModelProviderContextInput /MOutput /MAccess
Hunyuan Hy3FrontierCost-efficient frontier reasoning and agent work, positioned against other top open models.
TTencent (Hunyuan)
256K
$0.0825
$0.33
API (model id: hy3) via Tencent Cloud TokenHub, plus open weights on Hugging Face and GitHub. Also in Yuanbao, CodeBuddy, and ima.
MiniMax-M3FrontierFrontier coding, agentic execution, and multimodal reasoning at low cost.
MMiniMax
1M
$0.30
$1.20
API (model id: MiniMax-M3) via platform.minimax.io, plus open weights on Hugging Face.
Muse Spark 1.3FrontierAgentic and coding work on Meta's API.
Meta
1M
$1.25
$4.25
API
Muse Spark 1.2FrontierTeams already built on Muse Spark 1.2.
Meta
1M
$1.25
$4.25
API
GLM 5.3FrontierCoding agents and long tasks.
ZZhipu AI (GLM)
1.25M
$1.40
$4.40
API + open weights
GLM-5.2FrontierFrontier agentic coding and long-context reasoning at a fraction of closed-model cost.
ZZhipu AI (GLM)
1M
$1.40
$4.40
API (model id: glm-5.2) via Z.ai, plus the GLM Coding Plan and open weights on Hugging Face.
Gemini 3.1 ProFrontierComplex analysis over very long documents. Rates rise above 200K tokens.
Google
2M
$2.00
$12.00
API + Gemini app
Kimi K3FrontierComplex multi-step agents, deep research, and large-codebase work where the top tier matters.
KMoonshot AI (Kimi)
1M
$3.00
$15.00
API (model id: kimi-k3) at api.moonshot.ai, plus kimi.com and the Kimi apps. Open weights planned on Hugging Face.
GPT-5.6 SolFrontierTeams tackling the hardest agentic and coding work.
OpenAI
1M
$4.00
$20.00
API + ChatGPT
GPT-6 AstraFrontierThe hardest reasoning, coding and long agentic tasks.
OpenAI
1.05M
$10.00
$50.00
API
Mistral Large 3FlagshipTeams wanting a top open weight generalist with a hosted option.
Mistral AI
256K
$0.50
$1.50
API + open weights
Qwen3-MaxFlagshipThe hardest reasoning and long context workloads on Model Studio.
Alibaba (Qwen)
256K
$0.78
$3.90
API
Qwen3.8 MaxFlagshipThe hardest reasoning and coding on Alibaba Cloud.
Alibaba (Qwen)
1M
$2.00
$6.00
API
Claude Opus 5.5FlagshipAgentic coding and enterprise work at a lower price than Opus 5.
Anthropic
1M
$4.00
$20.00
API
Claude Opus 5FlagshipTeams already built on Opus 5.
Anthropic
1M
$5.00
$25.00
API
Claude Opus 4.8FlagshipComplex agentic coding and enterprise work.
Anthropic
1M
$5.00
$25.00
API + Claude apps
MiniMax-M1ReasoningLong-context reasoning and software tasks that need a big open thinking model.
MMiniMax
1M
$0.40
$2.20
API (model id: MiniMax-M1) via platform.minimax.io, plus open weights on GitHub and Hugging Face (Apache 2.0).
Grok 4.20ReasoningMulti-agent and long-horizon jobs.
xAI
2M
$1.25
$2.50
API
GLM-5.1ReasoningReasoning and coding work on a recent GLM-5 checkpoint.
ZZhipu AI (GLM)
1M
$1.40
$4.40
API (model id: glm-5.1) via Z.ai, plus open weights on Hugging Face.
o3ReasoningMath, logic, and analysis that need deeper thinking.
OpenAI
200K
$2.00
$8.00
API + ChatGPT
GPT-5.5 ProReasoningResearch-grade analysis where accuracy beats cost.
OpenAI
1M
$30.00
$180.00
API + ChatGPT
Command A ReasoningReasoningAgentic workflows that need deliberate reasoning.
CCohere
256K
API
Hunyuan-T1ReasoningDeliberate reasoning tasks in math, code, and analysis.
TTencent (Hunyuan)
64K
API (model id: hunyuan-t1) via Tencent Cloud.
GLM-4.6VBalancedMultimodal reasoning and document or UI understanding with tool use.
ZZhipu AI (GLM)
128K
$0.30
$0.90
API (model id: glm-4.6v) via Z.ai, plus open weights on Hugging Face.

Common questions

What is a reasoning model?

A reasoning model is tuned to work through a problem in explicit steps before giving a final answer, which helps on maths, logic, planning, and hard coding. It usually costs more and runs slower than a general model. Examples Vikshy tracks include Hunyuan Hy3 (Tencent (Hunyuan)), MiniMax-M3 (MiniMax), Muse Spark 1.3 (Meta).

← Back to the full model catalogue