Reasoning Rankings
ZICQ LLM rankings update daily across intelligence, coding, price, and context for selection and comparison.
Last updated: 39 min ago (2026-11-21 05:34)Sources and methodology
声明支持 reasoning 推理 effort 的模型,按 Intelligence 指数排序。
2. Benchmarks:Intelligence, coding, and agentic indices are sourced directly from Artificial Analysis (artificialanalysis.ai) official benchmark indices.
3. Overall Score Weights:Overall Score = 60% × Intelligence Index + 25% × Coding Index + 15% × Agentic Index. Models without benchmark data are ranked at the end.
4. Referral links:"Try it" buttons marked "Referral" are referral links; ZICQ may earn a commission. Referral relationships never affect rankings or scores.
Statistics for this selection
⚡ Side-by-Side Model PK & Token Cost Calculator Click "+ Compare" on any model below to swap · Live monthly cost calculator · Shareable comparison URL
| Metric | Anthropic: Claude Fable 5.2 Anthropic | Anthropic: Claude Fable 5.2 (batch) Anthropic | OpenAI: GPT-6.2 Astra OpenAI |
|---|---|---|---|
| Intelligence / Coding / Agentic | 64.6 / 0.0 / 0.0 | 64.6 / 0.0 / 0.0 | 62.1 / 0.0 / 0.0 |
| Context Window | 1M tokens | 1M tokens | 1.1M tokens |
| Input / Output Price (1M) | $8.00 / $40.00 | $4.00 / $20.00 | $8.00 / $40.00 |
| Est. Monthly Cost (30d) | Adjust inputs to calculate | Adjust inputs to calculate | Adjust inputs to calculate |
-
#381
Devstral 2 is a state-of-the-art open-source model by Mistral AI specializing in agentic coding. It is a 123B-parameter dense transformer model supporting a 256K context window. Devstral 2 supports exploring...
Tools Docs StructuredContext 262K Output ≤ 209K$0.40/1M $2.00/1M out cache $0.04/1MSolid Reasoning -
#382
The relace-search model uses 4-12 `view_file` and `grep` tools in parallel to explore a codebase and return relevant files to the user request. In contrast to RAG, relace-search performs agentic...
Tools StructuredContext 256K Output ≤ 128K$1.00/1M $3.00/1M outSolid Reasoning -
#383
The largest model in the Ministral 3 family, Ministral 3 14B offers frontier capabilities and performance comparable to its larger Mistral Small 3.2 24B counterpart. A powerful and efficient language...
Tools Vision StructuredContext 262K Output ≤ 209K$0.20/1M $0.20/1M out cache $0.02/1MSolid Reasoning -
#384
A balanced model in the Ministral 3 family, Ministral 3 8B is a powerful, efficient tiny language model with vision capabilities.
Tools Vision StructuredContext 262K Output ≤ 209K$0.15/1M $0.15/1M out cache $0.01/1MSolid Reasoning -
#385
The smallest model in the Ministral 3 family, Ministral 3 3B is a powerful, efficient tiny language model with vision capabilities.
Tools Vision StructuredContext 131K Output ≤ 104K$0.10/1M $0.10/1M out cache $0.01/1MSolid Reasoning -
#386
Mistral Large 3 2512 is Mistral’s most capable model to date, featuring a sparse mixture-of-experts architecture with 41B active parameters (675B total), and released under the Apache 2.0 license.
Tools Vision Docs StructuredContext 262K Output ≤ 209K$0.50/1M $1.50/1M out cache $0.05/1MSolid Reasoning -
#387
Amazon Nova Premier is the most capable of Amazon’s multimodal models for complex reasoning tasks and for use as the best teacher for distilling custom models.
Tools VisionContext 1M Output ≤ 32K$2.50/1M $12.50/1M out cache $0.62/1MSolid Reasoning -
#388
Voxtral Small is an enhancement of Mistral Small 3, incorporating state-of-the-art audio input capabilities while retaining best-in-class text performance. It excels at speech transcription, translation and audio understanding. Input audio...
Tools Audio Docs StructuredContext 32K Output ≤ 26K$0.10/1M $0.30/1M out cache $0.01/1MSolid Reasoning -
#389
Qwen3-VL-8B-Instruct is a multimodal vision-language model from the Qwen3-VL series, built for high-fidelity understanding and reasoning across text, images, and video. It features improved multimodal fusion with Interleaved-MRoPE for long-horizon...
Tools Vision StructuredContext 262K Output ≤ 235K$0.25/1M $0.75/1M out cache $0.12/1MSolid Reasoning -
#390
Gemini 2.5 Flash Image, a.k.a. "Nano Banana," is now generally available. It is a state of the art image generation model with contextual understanding. It is capable of image generation,...
Vision Structured Image genContext 32K Output ≤ 8K$0.30/1M $2.50/1M out cache $0.03/1MSolid Reasoning -
#391
Qwen3-VL-30B-A3B-Instruct is a multimodal model that unifies strong text generation with visual understanding for images and videos. Its Instruct variant optimizes instruction-following for general multimodal tasks. It excels in perception...
Tools Vision StructuredContext 262K Output ≤ 16K$0.15/1M $0.60/1M outSolid Reasoning -
#392
Uncensored and creative writing model based on Mistral Small 3.2 24B with good recall, prompt adherence, and intelligence.
StructuredContext 131K Output ≤ 117K$0.30/1M $0.50/1M out cache $0.15/1MSolid Reasoning -
#393
Relace Apply 3 is a specialized code-patching LLM that merges AI-suggested edits straight into your source files. It can apply updates from GPT-4o, Claude, and others into your files at...
Context 256K Output ≤ 128K$0.85/1M $1.25/1M outSolid Reasoning -
#394
Qwen3-VL-235B-A22B Instruct is an open-weight multimodal model that unifies strong text generation with visual understanding across images and video. The Instruct model targets general vision-language use (VQA, document parsing, chart/table...
Tools Vision StructuredContext 262K Output ≤ 32K$0.21/1M $1.90/1M out cache $0.10/1MSolid Reasoning -
#395
Qwen3 Coder Flash is Alibaba's fast and cost efficient version of their proprietary Qwen3 Coder Plus. It is a powerful coding agent model specializing in autonomous programming via tool calling...
Tools StructuredContext 1M Output ≤ 65K$0.20/1M $0.97/1M out cache $0.04/1MSolid Reasoning -
#396
Qwen3-Next-80B-A3B-Instruct is an instruction-tuned chat model in the Qwen3-Next series optimized for fast, stable responses without “thinking” traces. It targets complex tasks across reasoning, code generation, knowledge QA, and multilingual...
Tools StructuredContext 262K Output ≤ 235K$0.10/1M $1.10/1M out cache $0.07/1MSolid Reasoning -
#397
Kimi K2 0905 is the September update of [Kimi K2 0711](moonshotai/kimi-k2). It is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32...
Tools StructuredContext 262K Output ≤ 98K$0.60/1M $2.50/1M outSolid Reasoning -
#398
Mistral Medium 3.1 is an updated version of Mistral Medium 3, which is a high-performance enterprise-grade language model designed to deliver frontier-level capabilities at significantly reduced operational cost. It balances...
Tools Vision Docs StructuredContext 131K Output ≤ 104K$0.44/1M $2.20/1M out cache $0.04/1MSolid Reasoning -
#399
Mistral's cutting-edge language model for coding released end of July 2025. Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM), code correction and test generation. [Blog Post](https://mistral.ai/news/codestral-25-08)
Tools Docs StructuredContext 256K Output ≤ 204K$0.30/1M $0.90/1M out cache $0.03/1MSolid Reasoning -
#400
Qwen3-Coder-30B-A3B-Instruct is a 30.5B parameter Mixture-of-Experts (MoE) model with 128 experts (8 active per forward pass), designed for advanced code generation, repository-scale understanding, and agentic tool use. Built on the...
Tools StructuredContext 262K Output ≤ 235K$0.15/1M $0.60/1M outSolid Reasoning