Which AI model for: Code locally on a 24 GB card?

1. Gemma 4 31B (Google) · 2. Nemotron 3 Nano 30B-A3B (NVIDIA) · 3. Qwen2.5-Coder 32B (Alibaba). Deterministic nAIvigate selection on verified, dated data.

Which AI model for: An assistant on a laptop without GPU?

1. Granite 4.1 8B (IBM) · 2. Gemma 3n E4B (Google) · 3. Llama 3.1 8B (Meta). Deterministic nAIvigate selection on verified, dated data.

Which AI model for: Code via API, normal budget?

1. Gemini 3.7 Flash (Google) · 2. Kimi K2.6 (Moonshot AI) · 3. Qwen3.8-Max (Alibaba). Deterministic nAIvigate selection on verified, dated data.

Which AI model for: Generate images online?

1. DALL-E 3 (OpenAI) · 2. FLUX.1 Pro (Black Forest Labs). Deterministic nAIvigate selection on verified, dated data.

Which AI model for: RAG on internal documents, EU hosting?

1. Gemini 3.8 Flash (Google) · 2. Mistral Large 3 (Mistral AI) — Best European · 3. Qwen3.7 Max (Alibaba). Deterministic nAIvigate selection on verified, dated data.

Which AI model for: RAG on sensitive data, on your servers?

1. Kimi K2.6 (Moonshot AI) · 2. Mistral Large 3 (Mistral AI) — Best European · 3. GLM-5 (Zhipu AI). Deterministic nAIvigate selection on verified, dated data.

Which AI model for: Agents and automation via API?

1. Muse Spark 1.1 (Meta) · 2. Mistral Large 3 (Mistral AI) — Best European · 3. Grok 4.5 (xAI). Deterministic nAIvigate selection on verified, dated data.

Which AI model for: Translation and multilingual, tight budget?

1. GLM-5 (Zhipu AI) — Top ranked · 2. Qwen3.7 Plus (Alibaba) — Best via API · 3. DeepSeek V3.2 (DeepSeek) — Best value. Deterministic nAIvigate selection on verified, dated data.

Which AI model for: High-volume customer service?

1. GLM-5 (Zhipu AI) · 2. Mistral Small 3 (Mistral AI) — Best European · 3. Qwen3.7 Plus (Alibaba). Deterministic nAIvigate selection on verified, dated data.

Which AI model for your need?

Three to six questions. Each answer narrows the list, live, on the comparator’s verified data. No model is picked by an AI: the ranking is deterministic and every reason is a fact.

Step 1

Who is this for?

The next questions adapt to your answer.

For me

Personal or dev use, on my machine or online

For my team or company

Internal data, hosting, compliance, volume

Typical paths

Nine common needs, already answered

Same rules as the Finder, computed on today’s data. Click a path to resume and adjust it.

Code locally on a 24 GB card

For meCodeOn my machinePC with 24 GB card

19 models pass these constraints

1
Gemma 4 31BGoogle
Arena Elo 1451 — top-ranked within your constraints · Runs in Q4 on your machine (21 GB)
Context 256k · apache-2.0
2
Nemotron 3 Nano 30B-A3BNVIDIA
Arena Elo 1314 — #3 within your constraints · Runs in Q4 on your machine (21 GB)
Context 1M · nvidia-open-model
3
Qwen2.5-Coder 32BAlibaba
Arena Elo 1270 — #4 within your constraints · Runs in Q4 on your machine (23 GB)
Context 128k · apache-2.0

Resume this path

An assistant on a laptop without GPU

For meChat, think, searchOn my machineLaptop, no dedicated GPU

30 models pass these constraints

1
Granite 4.1 8BIBM
Arena Elo 1306 — top-ranked within your constraints · Runs in Q4 on your machine (7 GB)
Context 512k · apache-2.0
2
Gemma 3n E4BGoogle
Arena Elo 1317 — #2 within your constraints · Runs in Q4 on your machine (7 GB)
Context 32k · gemma
3
Llama 3.1 8BMeta
Arena Elo 1211 — #4 within your constraints · Runs in Q4 on your machine (7 GB)
Context 128k · llama-3.1-community

Resume this path

Code via API, normal budget

For meCodeOnline, via APINormal

18 models pass these constraints

1
Gemini 3.7 FlashGoogle
Arena Elo 1491 — top-ranked within your constraints · Code-specialised
Context 1.0M · Output / M $3.75 · proprietary
2
Kimi K2.6Moonshot AI
Arena Elo 1461 — #2 within your constraints · Code-specialised
Context 262k · Output / M $4.00 · modified-mit
3
Qwen3.8-MaxAlibaba
Arena Elo 1480 — #3 within your constraints · Code-specialised
Context 1M · Output / M $6.00 · proprietary

Resume this path

Generate images online

For meGenerate imagesOnline, via APINo constraint

2 models pass these constraints

1
DALL-E 3OpenAI
No Arena score yet — ranked on popularity
Context 4k · Output / M $0.00 · proprietary
2
FLUX.1 ProBlack Forest Labs
No Arena score yet — ranked on popularity
Context 4k · Output / M $0.00 · proprietary

Resume this path

RAG on internal documents, EU hosting

For my team or companyAnalyse our documentsInternalWith an EU hostOne departmentNormal

99 models pass these constraints

1
Gemini 3.8 FlashGoogle
Arena Elo 1494 — top-ranked within your constraints · EU hosting option from the vendor
Context 1.0M · Output / M $3.75 · proprietary
2
Mistral Large 3Mistral AI· Best European
Arena Elo 1414 — #21 within your constraints · European vendor (Mistral AI)
Context 256k · Output / M $6.00 · apache-2.0
3
Qwen3.7 MaxAlibaba
Arena Elo 1474 — #7 within your constraints · 1M context — comfortable for RAG
Context 1M · Output / M $7.50 · proprietary

Resume this path

RAG on sensitive data, on your servers

For my team or companyAnalyse our documentsSensitive or regulatedOn our serversGPU serverThe whole company

68 models pass these constraints

1
Kimi K2.6Moonshot AI
Arena Elo 1461 — top-ranked within your constraints · Runs unquantised on your machine
Context 262k · Output / M $4.00 · modified-mit
2
Mistral Large 3Mistral AI· Best European
Arena Elo 1414 — #8 within your constraints · Runs unquantised on your machine (1487 GB)
Context 256k · Output / M $6.00 · apache-2.0
3
GLM-5Zhipu AI
Arena Elo 1458 — #2 within your constraints · Runs unquantised on your machine
Context 128k · Output / M $2.00 · apache-2.0

Resume this path

Agents and automation via API

For my team or companyAgents and automationPublic or anonymisedOnline, via APIOne departmentNormal

8 models pass these constraints

1
Muse Spark 1.1Meta
Arena Elo 1492 — top-ranked within your constraints · Built for tool calling and agents
Context 1.0M · Output / M $4.25 · proprietary
2
Mistral Large 3Mistral AI· Best European
Arena Elo 1414 — #6 within your constraints · Built for tool calling and agents
Context 256k · Output / M $6.00 · apache-2.0
3
Grok 4.5xAI
Arena Elo 1471 — #2 within your constraints · Built for tool calling and agents
Context 500k · Output / M $6.00 · proprietary

Resume this path

Translation and multilingual, tight budget

For my team or companyTranslation and multilingualInternalCloud is fineOne departmentTight

38 models pass these constraints

1
GLM-5Zhipu AI· Top ranked
Arena Elo 1458 — top-ranked within your constraints
Context 128k · Output / M $2.00 · apache-2.0
2
Qwen3.7 PlusAlibaba· Best via API
Arena Elo 1455 — #3 within your constraints · API hosted in China
Context 1M · Output / M $1.60 · proprietary
3
DeepSeek V3.2DeepSeek· Best value
Arena Elo 1425 — #4 within your constraints · 5× cheaper than #1 per token
Context 131k · Output / M $0.42 · mit

Resume this path

High-volume customer service

For my team or companyCustomer servicePublic or anonymisedOnline, via APIThe whole companyTight

10 models pass these constraints

1
GLM-5Zhipu AI
Arena Elo 1458 — top-ranked within your constraints
Context 128k · Output / M $2.00 · apache-2.0
2
Mistral Small 3Mistral AI· Best European
Arena Elo 1235 — #8 within your constraints · First-class French
Context 32k · Output / M $0.60 · apache-2.0
3
Qwen3.7 PlusAlibaba
Arena Elo 1455 — #2 within your constraints · API hosted in China
Context 1M · Output / M $1.60 · proprietary

Resume this path