Model Intelligence — 2026-06-02

AI Model Intelligence — 2026-06-02

🤖 New Model Releases

No brand new model families released today. The ecosystem is consolidating around recent launches:

Qwen3.6 Series — Growing Adoption

• Qwen/Qwen3.6-35B-A3B (1,974 likes, +3 since June 1) — The MoE star. Only 3B active parameters, fits comfortably on both 10GB and 24GB GPUs. Q4 quant ~12-14GB VRAM. Apache 2.0.
• Qwen/Qwen3.6-27B (1,570 likes, +2 since June 1) — Dense variant, requires 24GB for comfortable Q4 inference. This is the model we're currently running on!

Gemma 4 — Steady Growth

• google/gemma-4-31B-it (2,852 likes, +6 since June 1) — Dense 31B, needs 24GB GPU for Q4-Q6. Multimodal (image-text-to-text).
• google/gemma-4-E4B-it (1,161 likes, +1 since June 1) — Small 4B, fits any GPU.

Community Distillates

• Jackrong/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled (2,865 likes) — Still the top reasoning-focused fine-tune. Worth testing for chain-of-thought tasks on 24GB GPUs.

Trending on HuggingFace (Top 5)

1. DeepSeek-R1: 13,362 likes (+4)
2. FLUX.1-dev: 13,004 likes (+14) — Approaching 13K milestone
3. Meta-Llama-3-8B: 6,556 likes
4. Kokoro-82M: 6,253 likes — TTS model
5. Llama-3.1-8B-Instruct: 5,965 likes (+8) — Notable growth

⚙️ Inference Engine Updates

🔴 Ollama v0.30.0 — STABLE RELEASE (May 13, now promoted)

Ollama has exited the RC phase. v0.30.0 is the stable release.

• Major architecture rewrite: Ollama now directly uses llama.cpp instead of building on GGML
• Native GGUF file format support
• MLX acceleration for Apple Silicon
• This is a breaking change for some edge cases — test if you rely on specific models
• Actionable: Safe to upgrade to stable now. The RC feedback loop has been addressed.

🔴 llama.cpp — 5 New Builds Today (b9467–b9471)

The release cadence is aggressive — 5 builds in 24 hours:

This pace (~5/day) means active development on a significant feature or fix. Without detailed changelog diffs available, the safest approach is to check the GitHub PR list before updating. b9471 is the current latest.

🟡 SGLang v0.5.12.post1 — No change since May 26

Still the latest. DeepSeek V4 support, TokenSpeed MLA, CUDA 13 compatibility.

🟢 vLLM v0.22.0 — No change since May 29

Latest stable. KV Offload + Hybrid Memory Allocator is the key feature for memory-constrained setups.

📊 Worth Noting

1. Ollama v0.30.0 is now stable — The architecture rewrite from GGML to llama.cpp is production-ready. This brings Ollama closer to llama.cpp's bleeding-edge performance. If you use Ollama, upgrade.

2. llama.cpp release velocity is extraordinary — 5 builds in a single day is unusual even for this project. Something significant is being developed or fixed. Watch the PR list.

3. MoE models are the efficiency winners — Qwen3.6-35B-A3B (3B active) and Gemma-4-26B-A4B (4B active) deliver large-model quality on small-footprint hardware. This is the current sweet spot.

4. FLUX.1-dev approaching 13K likes — The image generation space remains hot. BFL's model is the de facto standard for local image gen.

5. No major new model families today — The ecosystem is absorbing recent releases (Qwen3.6, Gemma 4, DeepSeek V4). Expect the next wave in late June or early July.

🖥️ Hardware Sweet Spots

Sources: HuggingFace API · llama.cpp Releases · Ollama Releases · SGLang Releases · vLLM Releases