Filter candidates by latency and cost first, then run small A/B tests on critical tasks. Avoid production decisions from rank alone.
Text LLMs from OpenRouter listings; composite score is a context/price proxy (not a benchmark).
Source: OpenRouter
Metrics: context_length_desc + avg_price_asc + owner_diversity
Update cadence: daily
Fetched at: 2026-07-10T14:42:11.210Z
Data version: v20260710T144210Z
Data size: 20
Method: webx-ranking-v2
| Rank | Model | Listing score | Key metric | 1M tokens (avg) |
|---|---|---|---|---|
| 1 | Qwen: Qwen3 Coder 480B A35B (free) | 99.9 | 1.0M ctx | $0.00 |
| 2 | DeepSeek: DeepSeek V4 Flash | 99.5 | 1.0M ctx | $0.13 |
| 3 | NVIDIA: Nemotron 3 Ultra (free) | 98.8 | 1.0M ctx | $0.00 |
| 4 | Xiaomi: MiMo-V2.5-Pro | 97.4 | 1.0M ctx | $0.65 |
| 5 | Z.ai: GLM 5.2 | 96 | 1.0M ctx | $1.02 |
| 6 | MiniMax: MiniMax M1 | 93.7 | 1.0M ctx | $1.30 |
| 7 | Writer: Palmyra X5 | 86.7 | 1.0M ctx | $3.30 |
| 8 | Tencent: Hy3 (free) | 81.9 | 262k ctx | $0.00 |
| 9 | Poolside: Laguna XS 2.1 (free) | 81.9 | 262k ctx | $0.00 |
| 10 | inclusionAI: Ling-2.6-flash | 81.9 | 262k ctx | $0.02 |
| 11 | Cohere: North Mini Code (free) | 81.7 | 256k ctx | $0.00 |
| 12 | StepFun: Step 3.5 Flash | 81.1 | 262k ctx | $0.20 |
| 13 | Arcee AI: Trinity Large Thinking | 79.9 | 262k ctx | $0.52 |
| 14 | Kwaipilot: KAT-Coder-Pro V2 | 78.8 | 256k ctx | $0.75 |
| 15 | OpenAI: gpt-oss-120b (free) | 78.8 | 131k ctx | $0.00 |
| 16 | Meta: Llama 3.3 70B Instruct (free) | 78.8 | 131k ctx | $0.00 |
| 17 | Relace: Relace Apply 3 | 77.6 | 256k ctx | $1.05 |
| 18 | Morph: Morph V3 Large | 76.4 | 262k ctx | $1.40 |
| 19 | MoonshotAI: Kimi K2 Thinking | 75.8 | 262k ctx | $1.55 |
| 20 | AI21: Jamba Large 1.7 | 62 | 256k ctx | $5.00 |
List prices in this table use OpenRouter as the primary listing source. Official vendor rows are on the Token hub. Fetched: 2026-07-10T14:42:11.210Z.