Model Rankings

AI Model Quality Rankings

Every model on Frontière AI is ranked by two criteria: Best Value (capability-to-price ratio — bigger, smarter models at lower prices rank higher) and Best Performance (raw capability, verified score, multimodality, and context length). Sovereign EU models get a +30% boost in the value ranking.

Rankings

🥇

Best Value

🥇
GPT-OSS 20B (OpenAI)

20B · 6/6 ✓ · EU sovereign · €0.0056/100k in

🥈
Qwen3 Coder 30B

30.5B · 6/6 ✓ · EU sovereign · €0.0084/100k in

🥉
GPT-OSS 120B (OpenAI)

120B · 6/6 ✓ · EU sovereign · €0.0112/100k in

4
DeepSeek V4.1 Flash (552B MoE, 1M context)

552B · 6/6 ✓ · Fast access · €0.0182/100k in

5
Qwen3 32B

32.8B · 6/6 ✓ · EU sovereign · €0.0112/100k in

Best Performance

🥇
DeepSeek V4.1 Flash (552B MoE, 1M context)

552B · 6/6 ✓ · Fast access · €0.0182/100k in

🥈
Kimi K3 (2.8T parameters, 1M context)

2.8T · 5/6 ✓ · Fast access · €0.3654/100k in

🥉
GLM-5.2 (Zhipu/Z.ai)

753B · 6/6 ✓ · EU sovereign · €0.252/100k in

4
Qwen3.5 397B (multimodal)

403B · 6/6 ✓ · EU sovereign · €0.084/100k in

5
DeepSeek V4 Flash 0731 (1M context)

158B · 6/6 ✓ · EU sovereign · €0.056/100k in

Full catalog

#ModelSizePrice (input)JurisdictionVerified Score
1GPT-OSS 20B (OpenAI)20B€0.0056EU sovereign6/6
2Qwen3 Coder 30B30.5B€0.0084EU sovereign6/6
3GPT-OSS 120B (OpenAI)120B€0.0112EU sovereign6/6
4DeepSeek V4.1 Flash (552B MoE, 1M context)552B€0.0182Fast access6/6
5Qwen3 32B32.8B€0.0112EU sovereign6/6
6Mistral Small 3.2 24B24B€0.0126EU sovereign6/6
7Gemma 3 27B (Google)27B€0.0098Fast access0/6
8Qwen3.5 9B (fast, budget)9.65B€0.014EU sovereign6/6
9DeepSeek R1 (Distill Qwen 14B)14B€0.0098Fast access0/6
10DeepSeek R1 (Distill Qwen 32B)32B€0.0154Fast access0/6
11DeepSeek V4 Flash 0731 (1M context)158B€0.056EU sovereign6/6
12Qwen3.5 397B (multimodal)403B€0.084EU sovereign6/6
13Qwen3.8 27B (multimodal)27.8B€0.056EU sovereign5/6
14Qwen3.6 27B (multimodal)27.8B€0.0574EU sovereign6/6
15DeepSeek R1 (671B MoE)671B€0.0854Fast access0/6
16Llama 3.3 70B70B€0.0898EU sovereign5/6
17Qwen3 235B (instruct)235B€0.105EU sovereign0/6
18Qwen2.5-VL 72B (vision)73.4B€0.1232EU sovereign5/6
19DeepSeek V4 Pro (345B MoE)345B€0.1064Fast access0/6
20GLM-5.2 (Zhipu/Z.ai)753B€0.252EU sovereign6/6
21Qwen3.8 Max (2.4T, 1M context)2.4T€0.2422Fast access0/6
22Kimi K3 (2.8T parameters, 1M context)2.8T€0.3654Fast access5/6
23Muse Spark 1.1 (1M context)?€0.1512Fast access0/6
24Muse Spark 1.2 (1M context)?€0.1512Fast access0/6
25GLM-5.3 Flash (Zhipu/Z.ai)?€0.0098Fast access0/6
26Qwen3.8 Flash (Qwen)?€0.0182Fast access0/6
27GLM-5.3 (Zhipu/Z.ai)?€0.1694Fast access0/6

Ranking methodology

Two rankings: Best Value computes a capability-to-price score (log-scale parameters × sovereignty +30% × context bonus × multimodality × verified score) divided by input price. Best Performance ranks by raw capability (parameters × verified criteria × features). The goal: help you pick the right model for your use case, not just the cheapest one.

Frequently asked questions

How are models ranked on Frontière AI?
Two rankings: Best Value scores capability-to-price ratio (bigger, smarter models at lower prices rank higher). Best Performance ranks by raw capability, verified score, and features like multimodality and context length. Sovereign EU models get a +30% value boost.
What is the Frontière Verified score?
The Frontière Verified score measures 6 criteria: license verified, provenance verified, security analyzed, GDPR assessed, benchmarked, and API available. A model with 6/6 passes all checks.
Why are sovereign models ranked higher?
Frontière AI prioritizes EU data sovereignty. Models hosted on sovereign providers (OVHcloud, Scaleway, self-hosted) get a +30% value score boost because EU data residency matters for GDPR and AI Act compliance.

Ready to test verified models?

Create an account