Two synthetic hosts, Wren and Ash, on the tracker's current state. The script was checked claim by claim against the tracker page; every claim carries a verbatim quote.
Transcript
Wren Older cloud AI models are being systematically removed.
Ash Right. OpenAI, Anthropic, and Adobe all have 2026 deprecation schedules.
Ash That includes GPT-4.1, Claude Sonnet 4, and Gemini variants.
Wren So the push is to newer reasoning models or open-weights.
Ash Exactly. But for local 48GB hardware, the stack is stable.
Ash No new models fit that box. gpt-oss-20b and Qwen3-Coder-30B are still the picks.
Wren What about the big vendor claims this month?
Ash NVIDIA claims up to 1.9x higher throughput with llama.cpp and vLLM.
Ash But that's a vendor claim. New TTS models like CosyVoice2-0.5B are only cited secondhand.
Wren Unverified?
Ash Secondary or vendor-claim. The tracker's verdict: local tier settled, frontier tier is a price war.
Wren So the action is migrating off the deprecated cloud models.
Ash That's the data. See the full, machine-maintained tracker at DreamLab Research.