# Qwen3-Coder-30B-A3B-Instruct vs Recursive Flint > Qwen3-Coder-30B-A3B-Instruct (~22.3 GB at Q4) vs Recursive Flint (~28.7 GB): memory at Q4/Q8/FP16, context cache, cheapest GPU and Mac, and est. tok/s side by side. Source: https://whatcanirunai.vercel.app/compare/qwen3-coder-30b-a3b-instruct-vs-flint/ # Qwen3-Coder-30B-A3B-Instruct vs Recursive Flint: VRAM and speed for running locally Side by side: how much memory each model needs at Q4, Q8 and FP16, how the KV cache grows with context, the cheapest GPU and Mac that run each, and estimated speed on 11 common machines. **Qwen3-Coder-30B-A3B-Instruct** is lighter: about 22.3 GB at Q4 against 28.7 GB for Recursive Flint, though both run on 16 of the devices we track. **Recursive Flint** should be faster on the same hardware, since it reads 2.5B parameters per token against 3.0B. Which one gives better answers depends on your task, so check benchmarks for that; this page covers what it takes to run them. __Qwen3-Coder-30B-A3B-Instruct vs Recursive Flint: specs and memory__ | [Qwen3-Coder-30B-A3B-Instruct](https://whatcanirunai.vercel.app/models/qwen3-coder-30b-a3b-instruct/) | [Recursive Flint](https://whatcanirunai.vercel.app/models/flint/) | | | --------------------------------------------------------------------- | ------------------------------------------------------------ | ------------------------------------------------------------------------------ | | Maker | Qwen | RecursiveSI | | License | apache-2.0 | apache-2.0 | | Type | Mixture of experts | Mixture of experts | | Total parameters | 30.5B | 39.8B | | Active per token | 3.0B | 2.5B | | Memory at Q4\_K\_M | **22.3 GB** | **28.7 GB** | | Memory at Q8\_0 | **38.3 GB** | **49.6 GB** | | Memory at FP16 | **71.2 GB** | **92.5 GB** | | KV cache growth | 98 MB per 1k tokens, FP16 | 100 MB per 1k tokens, FP16 | | Memory at Q4, 32k context | 25.1 GB 3.2 GB cache | 29.6 GB 1.0 GB cache | | Cheapest GPU at Q4 | [Intel Arc Pro B70 32GB](https://whatcanirunai.vercel.app/hardware/arc-pro-b70/) $949 | [Intel Arc Pro B70 32GB](https://whatcanirunai.vercel.app/hardware/arc-pro-b70/) $949 | | Cheapest Mac / mini PC at Q4 | [Mac mini M5 Pro 64GB](https://whatcanirunai.vercel.app/hardware/mac-mini-m5-pro-64/) $3,199 | [Mac mini M5 Pro 64GB](https://whatcanirunai.vercel.app/hardware/mac-mini-m5-pro-64/) $3,199 | | Cheapest GPU at Q8 | [NVIDIA RTX 6000 Ada 48GB](https://whatcanirunai.vercel.app/hardware/rtx-6000-ada/) $6,799 | [NVIDIA RTX Pro 6000 Blackwell 96GB](https://whatcanirunai.vercel.app/hardware/rtx-pro-6000-blackwell/) $8,565 | | Tracked devices that run it at Q4 | 16 of 82 | 16 of 82 | ## Estimated speed on the same hardware Decode tokens per second at Q4\_K\_M from memory bandwidth ([how we estimate](https://whatcanirunai.vercel.app/method/#speed)). MoE figures are upper bounds ("up to"). Estimates, not benchmarks. __Estimated tokens per second, Qwen3-Coder-30B-A3B-Instruct vs Recursive Flint__ | Device | Qwen3-Coder-30B-A3B-Instruct | Recursive Flint | | --------------------------------------------------------------------------------------------------------------- | ---------------------------- | --------------- | | [GeForce RTX 3060 12GB](https://whatcanirunai.vercel.app/hardware/rtx-3060-12gb/)12 GB usable · 360 GB/s · $329 | doesn’t fit | doesn’t fit | | [GeForce RTX 4060 Ti 16GB](https://whatcanirunai.vercel.app/hardware/rtx-4060-ti-16gb/)16 GB usable · 288 GB/s · $499 | doesn’t fit | doesn’t fit | | [GeForce RTX 5080](https://whatcanirunai.vercel.app/hardware/rtx-5080/)16 GB usable · 960 GB/s · $999 | doesn’t fit | doesn’t fit | | [GeForce RTX 3090](https://whatcanirunai.vercel.app/hardware/rtx-3090/)24 GB usable · 936 GB/s · $1,499 | tight fit | doesn’t fit | | [Radeon RX 7900 XTX](https://whatcanirunai.vercel.app/hardware/rx-7900-xtx/)24 GB usable · 960 GB/s · $999 | tight fit | doesn’t fit | | [GeForce RTX 5090](https://whatcanirunai.vercel.app/hardware/rtx-5090/)32 GB usable · 1792 GB/s · $1,999 | up to 837 | up to 1005 | | [Mac mini M6 24GB](https://whatcanirunai.vercel.app/hardware/mac-mini-m6-24/)18 GB usable · 170 GB/s · $1,299 | doesn’t fit | doesn’t fit | | [Mac mini M6 32GB](https://whatcanirunai.vercel.app/hardware/mac-mini-m6-32/)24 GB usable · 170 GB/s · $1,799 | tight fit | doesn’t fit | | [Mac mini M5 Pro 64GB](https://whatcanirunai.vercel.app/hardware/mac-mini-m5-pro-64/)48 GB usable · 307 GB/s · $3,199 | up to 135 | up to 162 | | [Framework Desktop (Ryzen AI Max+ 395) 128GB](https://whatcanirunai.vercel.app/hardware/framework-desktop-128/)96 GB usable · 256 GB/s · $3,449 | up to 113 | up to 135 | | [MacBook Pro M5 Max 128GB](https://whatcanirunai.vercel.app/hardware/macbook-pro-m5-max-128/)96 GB usable · 614 GB/s · $6,699 | up to 270 | up to 324 | ## Quick answers Which needs less VRAM, Qwen3-Coder-30B-A3B-Instruct or Recursive Flint? Qwen3-Coder-30B-A3B-Instruct: about 22.3 GB at Q4\_K\_M, against 28.7 GB for Recursive Flint. At Q8\_0 it's 38.3 GB vs 49.6 GB. Which is faster locally, Qwen3-Coder-30B-A3B-Instruct or Recursive Flint? Decode speed is limited by how many bytes are read per token. Recursive Flint reads 2.5B parameters per token and Qwen3-Coder-30B-A3B-Instruct reads 3.0B, so Recursive Flint should be faster on the same machine. Our estimate on the GeForce RTX 5090 at Q4 is up to 1005 tok/s vs up to 837 tok/s. These are estimates, not benchmarks. What is the cheapest GPU for Qwen3-Coder-30B-A3B-Instruct and Recursive Flint? At Q4: Intel Arc Pro B70 32GB ($949) for Qwen3-Coder-30B-A3B-Instruct; Intel Arc Pro B70 32GB ($949) for Recursive Flint. Prices are dated list-price snapshots. Can a Mac run Qwen3-Coder-30B-A3B-Instruct and Recursive Flint? The cheapest Mac or mini PC we track for Qwen3-Coder-30B-A3B-Instruct is the Mac mini M5 Pro 64GB ($3,199). For Recursive Flint it's the Mac mini M5 Pro 64GB ($3,199). Which is better, Qwen3-Coder-30B-A3B-Instruct or Recursive Flint? We don't score answer quality. This comparison covers memory, context cache, hardware and speed. Pick the one that fits your hardware, then test both on your own prompts. [All comparisons](https://whatcanirunai.vercel.app/compare/) · [MoE vs dense](https://whatcanirunai.vercel.app/moe-vs-dense/) · [Q4 vs Q8 vs FP16](https://whatcanirunai.vercel.app/quantization/) · [Context length and VRAM](https://whatcanirunai.vercel.app/context-length/) · [VRAM calculator](https://whatcanirunai.vercel.app/vram-calculator/) --- Agent guide: https://whatcanirunai.vercel.app/llms.txt · JSON API: https://whatcanirunai.vercel.app/developers/