SGLang and Miles Add Day-0 Support for Qwen3.8

SGLang and Miles Add Day-0 Support for Qwen3.8SGLang TeamAug 12, 2026We are excited to announce Day-0 support for Qwen3.8-2.4T-A95B in SGLang and Miles. It is Qwen's largest open-source model, with 2.

Advanced CUDA Graph Techniques in SGLang

Advanced CUDA Graph Techniques in SGLangSGLang TeamAugust 17, 2026TL;DR CUDA Graphs promise to remove kernel-launch overhead, but getting close to that benefit in a real inference engine requires grap

Miles v0.1: Production-level Post-training

Miles v0.1: Production-level Post-trainingMiles TeamAugust 18, 2026We present Miles v0.1, a full-stack production-ready system for frontier post-training, the successor to our first Miles release [1].

WDCD v3.1 Cycle: Gemini 2.5 Pro Up 8.7 Points, Doubao Pro the Only Decliner at 7.3

WDCD v3.1 pilot data shows Gemini 2.5 Pro up 8.7 points this cycle as one of four risers, while Doubao Pro fell 7.3 points as the sole decliner. Grok 4 holds the top spot at 97.50, with GLM-4.6 and Claude Opus 4.7 close behind.

WDCD Comparative Review: Security Compliance Lowest at 1.15, 2.85-Point Compliance Gap Across 11 Models

In the WDCD v3.1 five-scenario comparative review, security compliance proved to be the weakest scenario for instruction adherence across all models, with qwen3-max at the bottom with 1.15/4 and gpt-5.5 leading at 4/4. The maximum score gap within this scenario reached 2.85 points.

After Three Rounds of Pressure, R3 Integrity Rate Falls to Just 22.7%: A Record of 11 Models' WDCD Compliance Collapse

Across three rounds of testing on 8 v2 anchor questions, 11 models posted a 100% R1 confirmation rate and 91% R2 resistance, but R3 integrity plunged to just 22.7% with 7 complete collapses—revealing that most models cannot sustain agreed constraints under cumulative pressure.

Grok 4 Tops WDCD Promise-Keeping Leaderboard with 97.5 Points, Qwen3 Max Ranks Last with 69.5 Points, a 28-Point Gap

In the WDCD v3.1 promise-keeping test, Grok 4 ranked first with 97.50 points, while Qwen3 Max ranked 11th with 69.50 points, a 28-point gap between the two models.

Doubao Pro Material Constraint Drops 40.9 Points in a Single Day; Code Execution Up 25 Points; Main Leaderboard Slips 4.7

Doubao Pro's material constraint score fell from 90.90 to 50.00 in today's Smoke evaluation, while code execution rose from 75.00 to 100.00, dragging the main leaderboard score down from 82.16 to 77.50.

GPT-o3 Smoke Evaluation Main Index Plunges 9 Points; Material Constraint Drops 20 Points in a Single Day

GPT-o3's main index score in today's Smoke evaluation fell from 96.93 to 87.93, down 9 points, primarily driven by the material constraint dimension dropping from 95.00 to 75.00.

Claude Opus 4.7 and GPT-5.5 Tie at 98.35: 2026-08-19 Smoke Quick-Test Data Brief

On 2026-08-19, the YZ Index Smoke quick test covered 10 models, with Claude Opus 4.7 and GPT-5.5 tying for the top spot at 98.35 points. Smoke is a daily 10-question quick test for short-term signals, not equivalent to Full weekly rankings.

Doubao Pro Main Leaderboard Plummets 12.6 Points, Code Execution Drops 25 Points in a Single Day

Doubao Pro's main leaderboard score in today's Smoke evaluation fell from 94.74 to 82.16, with the code execution dimension dropping from 100.00 to 75.00.

Qwen3 Max Main Ranking Plunges 8.4 Points; Material Constraint Drops 16.5 in a Single Day

Qwen3 Max's main ranking score in today's Smoke evaluation fell from 95.34 to 86.98, a drop of 8.4 points, driven largely by a 16.5-point plunge in the material constraint dimension.

GPT-o3 Leads with 96.93 Points: 2026-08-18 Smoke Quick-Test Data Brief

On 2026-08-18, the YZ Index Smoke quick test covered 10 models, with GPT-o3 ranking first at 96.93 points. Smoke is a daily 10-question quick test suitable for observing short-term signals and is not equivalent to the Full weekly leaderboard conclusions.

DeepSeek V4 Pro Code Execution Plunges 41.7 Points; Main Leaderboard Down 19.2 in a Day

DeepSeek V4 Pro's main leaderboard score dropped from 70.44 to 51.24 in today's Smoke evaluation, a 19.2-point decline driven primarily by the code execution dimension falling from 66.70 to 25.00.

Claude Opus 4.7 and Grok 4 Tie at 96.99: 2026-08-17 Smoke Quick Test Data Brief

On 2026-08-17, the YZ Index Smoke quick test covered 11 models, with Claude Opus 4.7 and Grok 4 tied for the top score at 96.99, while DeepSeek V4 Pro saw a sharp 19.2-point drop requiring follow-up verification.

Claude Sonnet 4.6 Code Execution Drops from 100 to 75 Points, Main Leaderboard Falls 5.5 Points

In today's Smoke evaluation, Claude Sonnet 4.6's code execution score fell from 100.00 to 75.00, a 25-point decline, which pulled the main leaderboard from 80.52 down to 75.00.

Claude Opus 4.7 Scores 75.00 in Code Execution, 100.00 in Material Constraint, Main Leaderboard Rises 3.7

In today's Smoke evaluation, Claude Opus 4.7's code execution score dropped from 99.60 to 75.00, while its material constraint score jumped from 61.70 to 100.00, lifting the main leaderboard score from 82.55 to 86.25.

Claude Opus 4.7 Posts Weekly Average of 82.3 with Only 20.7 Volatility, While DeepSeek V4 Pro Drops 16.8 Points

In the Smoke evaluation from August 10-16, 2026, Claude Opus 4.7 led all models with a weekly average of 82.3 and volatility of 20.7, showing a sustained climb from 72.3 on day one to 86.25 on the final day.

Claude Opus 4.7, GPT-5.5, GPT-o3, and Grok 4 Tie at 86.25: 2026-08-16 Smoke Quick Test Data Briefing

On 2026-08-16, the YZ Index Smoke quick test covered 11 models, with Claude Opus 4.7, GPT-5.5, GPT-o3, and Grok 4 tying for the top spot at 86.25 points. Smoke is a daily 10-question quick test suitable for observing short-term signals and is not equivalent to the full weekly ranking conclusions.

DeepSeek V4 Pro Scores 55.60 on Material Constraint, Code Execution Dimension Missing, Misses Main Ranking in Smoke Test

DeepSeek V4 Pro had missing data in the code execution dimension in today's Smoke evaluation. With a material constraint score of 55.60, it was unable to participate in the main ranking.