Claude Opus 5.5, GPT-6 Sol, GPT-6.1 Sol, Claude Haiku 5.5, Gemini 4 Argon, Grok 4.7, DeepSeek V4.1 Flash, and Qwen3.8 Max.
| Rank | Release Date | Model & Lab | Category | Architecture | Score | Action |
|---|---|---|---|---|---|---|
| #1 | Oct 7, 2026 |
Claude Haiku 5.5
Anthropic
|
Frontier Models & Flagships | High-Throughput Sub-Agent Transformer Β· Sub-Second Token Latency | 102.3 | Try β |
| #6 | Oct 6, 2026 |
Mistral Large 4 Preview
Mistral AI
|
Frontier Models & Flagships | Dense Multimodal Sovereign Transformer Β· Weights Planned Oct 27 | 92.0 | Try β |
| #12 | Oct 2, 2026 |
AstaBrief-8B
Allen Institute for AI
|
Frontier Models & Flagships | Open Scholarly Synthesis & Scientific CoT Reasoning Transformer Β· 8B Params | 84.9 | Try β |
| #14 | Oct 1, 2026 |
Clef-Flash
Cloudflare
|
Frontier Models & Flagships | Sub-10ms Visual Embedding & Bounding Box Model on Cloudflare Edge | 75.9 | Try β |
| #20 | Oct 1, 2026 |
Pareto 26.10 Preview
Unbiased
|
Frontier Models & Flagships | Dual-Phase Mathematical Search & Constraint Optimization Reasoner | 72.7 | Try β |
| #21 | Sep 30, 2026 |
Gemini 4 Argon
Google DeepMind
|
Frontier Models & Flagships | Next-Gen Argon Cluster Architecture Β· 2M Native Context Β· Multimodal Native | 74.8 | Try β |
| #40 | Sep 22, 2026 |
GPT-6 Luna
OpenAI
|
Frontier Models & Flagships | High-Throughput Cost-Efficient GPT-6 Tier ($0.40 / $1.60 per MTok) | 51.9 | Try β |
| #41 | Sep 22, 2026 |
Claude Opus 5.5
Anthropic
|
Frontier Models & Flagships | Highest Intelligence Frontier Model Β· SOTA LMSYS Arena Elo 1420 | 52.4 | Try β |
| #44 | Sep 21, 2026 |
Grok 4.7
xAI
|
Frontier Models & Flagships | Colossus-Trained Frontier MoE with Native Real-Time Grounding & Vision | 49.7 | Try β |
| #46 | Sep 18, 2026 |
Qwen3.8-LiveTranslate
Alibaba / Qwen
|
Frontier Models & Flagships | Simultaneous Streaming Speech & Text Machine Translation Model | 45.4 | Try β |
| #54 | Sep 10, 2026 |
GigaChat 3.5 432B A28B Reasoning
Sber / ai-sage
|
Frontier Models & Flagships | Sovereign Slavic Reasoning MoE Β· 432B Parameters (28B Active) | 29.9 | Try β |
| #55 | Sep 9, 2026 |
North Small Translate 1.0
Cohere Labs
|
Frontier Models & Flagships | Compact Specialized Polyglot Machine Translation Architecture | 28.3 | Try β |
| #59 | Sep 3, 2026 |
GPT-6 Astra
OpenAI
|
Frontier Models & Flagships | Ultra-Dense Frontier Intelligence Model with Integrated Multi-Step Search | 22.2 | Try β |
| #61 | Sep 2, 2026 |
Gemini 3.8 Flash
Google DeepMind
|
Frontier Models & Flagships | High-Speed Frontier Transformer Β· 1M Native Multimodal Context | 21.1 | Try β |
| #62 | Sep 2, 2026 |
Qwen3.8 Max 0902
Alibaba / Qwen
|
Frontier Models & Flagships | Alibaba Flagship Frontier MoE Snapshot Β· 128k Context Β· LMSYS Elo 1385 | 21.1 | Try β |
| #64 | Sep 1, 2026 |
Claude Mythos 5.1
Anthropic
|
Frontier Models & Flagships | Restricted Enterprise Cyber Defense & Formal Verification Engine | 20.3 | Try β |
| #76 | Aug 12, 2026 |
North Micro Vision Instruct
Cohere Labs
|
Frontier Models & Flagships | Compact Visual Representation & Diagram Parsing Transformer | 8.0 | Try β |
| #77 | Aug 12, 2026 |
Grok 4.6
xAI
|
Frontier Models & Flagships | Colossus-Trained Multimodal Reasoning Model with Real-Time Web Telemetry | 8.2 | Try β |
| #83 | Aug 2, 2026 |
Qwen3.8 Max
Alibaba / Qwen
|
Frontier Models & Flagships | Flagship Sovereign Frontier MoE Transformer with Long-Horizon Agent Capabilities | 5.2 | Try β |
| #88 | Jul 24, 2026 |
Claude Opus 5
Anthropic
|
Frontier Models & Flagships | Previous-Generation Flagship Intelligence Architecture Β· Elo 1395 | 3.5 | Try β |
| #91 | Jul 21, 2026 |
Gemini 3.6 Flash
Google
|
Frontier Models & Flagships | High-Throughput Multimodal Foundation Model with Sub-Second Response Times | 3.0 | Try β |
| #92 | Jul 21, 2026 |
Gemini 3.5 Flash-Lite
Google
|
Frontier Models & Flagships | Sub-Billion Scale Low-Power Edge Multimodal Engine | 3.0 | Try β |
| #96 | Jul 9, 2026 |
GPT-5.6 Sol
OpenAI
|
Frontier Models & Flagships | Previous-Generation Flagship Reasoning Transformer | 1.8 | Try β |
| #97 | Jul 9, 2026 |
GPT-5.6 Terra
OpenAI
|
Frontier Models & Flagships | Balanced Performance-Cost Enterprise Foundation Model | 1.8 | Try β |
| #98 | Jul 9, 2026 |
GPT-5.6 Luna
OpenAI
|
Frontier Models & Flagships | High-Speed Frontier Transformer for Low-Latency Applications | 1.8 | Try β |