Pokee-Isaac 28B Beats Meta and Qwen Rivals in Sub-30B Agent Benchmarks
Kyrannio · x · 2026-08-12
Pokee AI released a side-by-side comparison of its Pokee-Isaac 28B against Meta's Muse Glimmer 30B, Qwen3.6 27B, and Gemma 4 31B. In the 30B weight class, Pokee-Isaac leads across all tested agentic benchmarks: BFCL v4 (70.94), τ³-bench (66.67), Terminal-Bench 2.1 (65.12), and MCP-Atlas (74.59%).
It also features unique long-context capabilities, scoring 93.3% on RULER at 10M context and 0.5 on MRCR v2 at 1M tokens (where all others scored 0). Deployable on a single GPU, the local frontier for 30B models is heating up.
More from Models
- Microsoft's New Code Model Boosts Efficiency 25% at Quarter of the Cost — mustafasuleyman · 2026-08-12
- Users Report Grok Unreasonably Refusing Cutting-Edge Science Equations — Promptmethus · 2026-08-12
- Ling-3.0-flash Quantization Benchmarks: MoE Architecture Preserves Decode Speed — AcanthisittaOk1699 · 2026-08-12
- FLUX 3 Video Ranks #2 Globally, Free Access Limited Time — arena · 2026-08-12
- Researchers Spot Mysterious Gibberish from OpenAI Endpoint, Suspect Token Decoding Bug — jonasgeiping · 2026-08-12
- Muse Glimmer 30B Hits 25 tok/s In-Browser on M4 Max via Custom WebGPU Kernels — xenovatech · 2026-08-12